10 - Deprecated ``admin_up`` instance field. Instead, ``admin_state`` is
11 introduced, with 3 possible values -- ``up``, ``down`` and
13 - Replaced ``--disks`` option of ``gnt-instance replace-disks`` with a
14 more flexible ``--disk`` option. Now disk size and mode can be changed
16 - Removed deprecated ``QueryLocks`` LUXI request. Use
17 ``Query(what=QR_LOCK, ...)`` instead.
18 - The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
19 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
20 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
21 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
22 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
23 - RAPI client: ``CertificateError`` now derives from ``GanetiApiError``
24 - Deprecation warnings due to PyCrypto/paramiko import in
25 ``tools/setup-ssh`` have been silenced, as usually they are safe;
26 please make sure to run an up-to-date paramiko version
27 - The QA scripts now depend on Python 2.5 or above
33 *(Released Fri, 11 May 2012)*
35 A small bugfix release.
37 The main issues solved are on the topic of compatibility with newer LVM
40 - fixed parsing of ``lv_attr`` field
41 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
42 the ``--force`` flag is needed
44 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
45 to accept kernel 3.x too (was hardcoded to 2.6.*).
47 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
48 of hook results and that also made display of other errors suboptimal
49 was fixed; the code behaves now like 2.4 and earlier.
51 Another change in 2.5, the cleanup of the OS scripts environment, is too
52 aggressive: it removed even the ``PATH`` variable, which requires the OS
53 scripts to *always* need to export it. Since this is a bit too strict,
54 we now export a minimal PATH, the same that we export for hooks.
56 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
57 integrated into this release.
59 Finally, a few other miscellaneous changes were done (no new features,
60 just small improvements):
62 - Fix ``gnt-group --help`` display
63 - Fix hardcoded Xen kernel path
64 - Fix grow-disk handling of invalid units
65 - Update synopsis for ``gnt-cluster repair-disk-sizes``
66 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
72 *(Released Thu, 12 Apr 2012)*
74 Incompatible/important changes and bugfixes
75 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
77 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
78 resource's ``ip_check`` parameter changed from ``True`` to ``False``
79 to match the underlying LUXI interface.
80 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
81 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
82 not maintain backwards-compatibility as the underlying operation
83 changed in an incompatible way. The RAPI client can talk to old
84 servers, but it needs to be told so as the return value changed.
85 - When creating file-based instances via RAPI, the ``file_driver``
86 parameter no longer defaults to ``loop`` and must be specified.
87 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
89 - Support for the undocumented and deprecated RAPI instance creation
90 request format version 0 has been dropped. Use version 1, supported
91 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
92 - Pyparsing 1.4.6 or above is required, see :doc:`installation
93 documentation <install>`.
94 - The "cluster-verify" hooks are now executed per group by the
95 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
96 if you just run ``gnt-cluster verify``, which generates one opcode per
98 - The environment as passed to the OS scripts is cleared, and thus no
99 environment variables defined in the node daemon's environment will be
100 inherited by the scripts.
101 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
103 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
104 support operations involving multiple node groups.
105 - Offline nodes are ignored when failing over an instance.
106 - Support for KVM version 1.0, which changed the version reporting format
108 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
110 - ``Makefile`` is now compatible with Automake 1.11.2
111 - Includes all bugfixes made in the 2.4 series
116 - The ganeti-htools project has been merged into the ganeti-core source
117 tree and will be built as part of Ganeti (see :doc:`install-quick`).
118 - Implemented support for :doc:`shared storage <design-shared-storage>`.
119 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
120 GPT-style partition tables (requires `parted
121 <http://www.gnu.org/s/parted/>`_).
122 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
123 - Allowed adding tags on instance creation.
124 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
125 - Allow instances to be started in a paused state, enabling the user to
126 see the complete console output on boot using the console.
127 - Added new hypervisor flag to control default reboot behaviour
128 (``reboot_behavior``).
129 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
130 - Improved out-of-band management support:
132 - Added ``gnt-node health`` command reporting the health status of
134 - Added ``gnt-node power`` command to manage power status of nodes.
135 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
137 - Instance migration can fall back to failover if instance is not
139 - Filters can be used when listing nodes, instances, groups and locks;
140 see *ganeti(7)* manpage.
141 - Added post-execution status as variables to :doc:`hooks <hooks>`
143 - Instance tags are exported/imported together with the instance.
144 - When given an explicit job ID, ``gnt-job info`` will work for archived
146 - Jobs can define dependencies on other jobs (not yet supported via
147 RAPI or command line, but used by internal commands and usable via
150 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
153 - Instance failover is now available as a RAPI resource
154 (``/2/instances/[instance_name]/failover``).
155 - ``gnt-instance info`` defaults to static information if primary node
157 - Opcodes have a new ``comment`` attribute.
158 - Added basic SPICE support to KVM hypervisor.
159 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
161 Node group improvements
162 ~~~~~~~~~~~~~~~~~~~~~~~
164 - ``gnt-cluster verify`` has been modified to check groups separately,
165 thereby improving performance.
166 - Node group support has been added to ``gnt-cluster verify-disks``,
167 which now operates per node group.
168 - Watcher has been changed to work better with node groups.
170 - One process and state file per node group.
171 - Slow watcher in one group doesn't block other group's watcher.
173 - Added new command, ``gnt-group evacuate``, to move all instances in a
174 node group to other groups.
175 - Added ``gnt-instance change-group`` to move an instance to another
177 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
178 per-group operations.
179 - Node groups can be tagged.
180 - Some operations switch from an exclusive to a shared lock as soon as
182 - Instance's primary and secondary nodes' groups are now available as
183 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
184 and ``snodes.group.uuid``).
189 - Numerous updates to documentation and manpages.
191 - :doc:`RAPI <rapi>` documentation now has detailed parameter
193 - Some opcode/job results are now also documented, see :doc:`RAPI
196 - A lockset's internal lock is now also visible in lock monitor.
197 - Log messages from job queue workers now contain information about the
198 opcode they're processing.
199 - ``gnt-instance console`` no longer requires the instance lock.
200 - A short delay when waiting for job changes reduces the number of LUXI
201 requests significantly.
202 - DRBD metadata volumes are overwritten with zeros during disk creation.
203 - Out-of-band commands no longer acquire the cluster lock in exclusive
205 - ``devel/upload`` now uses correct permissions for directories.
211 *(Released Fri, 23 Mar 2012)*
213 This was the sixth release candidate of the 2.5 series.
219 *(Released Mon, 9 Jan 2012)*
221 This was the fifth release candidate of the 2.5 series.
227 *(Released Thu, 27 Oct 2011)*
229 This was the fourth release candidate of the 2.5 series.
235 *(Released Wed, 26 Oct 2011)*
237 This was the third release candidate of the 2.5 series.
243 *(Released Tue, 18 Oct 2011)*
245 This was the second release candidate of the 2.5 series.
251 *(Released Tue, 4 Oct 2011)*
253 This was the first release candidate of the 2.5 series.
259 *(Released Wed, 31 Aug 2011)*
261 This was the third beta release of the 2.5 series.
267 *(Released Mon, 22 Aug 2011)*
269 This was the second beta release of the 2.5 series.
275 *(Released Fri, 12 Aug 2011)*
277 This was the first beta release of the 2.5 series.
283 *(Released Thu, 27 Oct 2011)*
285 - Fixed bug when parsing command line parameter values ending in
287 - Fixed assertion error after unclean master shutdown
288 - Disable HTTP client pool for RPC, significantly reducing memory usage
290 - Fixed queue archive creation with wrong permissions
296 *(Released Tue, 23 Aug 2011)*
300 - Fixed documentation for importing with ``--src-dir`` option
301 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
302 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
308 *(Released Fri, 5 Aug 2011)*
310 Many bug-fixes and a few small features:
312 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
313 issues when you tried to add an instance with two MAC addresses in one
315 - KVM: fixed per-instance stored UID value
316 - KVM: configure bridged NICs at migration start
317 - KVM: Fix a bug where instance will not start with never KVM versions
319 - Added OS search path to ``gnt-cluster info``
320 - Fixed an issue with ``file_storage_dir`` where you were forced to
321 provide an absolute path, but the documentation states it is a
322 relative path, the documentation was right
323 - Added a new parameter to instance stop/start called ``--no-remember``
324 that will make the state change to not be remembered
325 - Implemented ``no_remember`` at RAPI level
326 - Improved the documentation
327 - Node evacuation: don't call IAllocator if node is already empty
328 - Fixed bug in DRBD8 replace disks on current nodes
329 - Fixed bug in recreate-disks for DRBD instances
330 - Moved assertion checking locks in ``gnt-instance replace-disks``
331 causing it to abort with not owning the right locks for some situation
332 - Job queue: Fixed potential race condition when cancelling queued jobs
333 - Fixed off-by-one bug in job serial generation
334 - ``gnt-node volumes``: Fix instance names
335 - Fixed aliases in bash completion
336 - Fixed a bug in reopening log files after being sent a SIGHUP
337 - Added a flag to burnin to allow specifying VCPU count
338 - Bugfixes to non-root Ganeti configuration
344 *(Released Thu, 12 May 2011)*
346 Many bug-fixes and a few new small features:
348 - Fixed a bug related to log opening failures
349 - Fixed a bug in instance listing with orphan instances
350 - Fixed a bug which prevented resetting the cluster-level node parameter
351 ``oob_program`` to the default
352 - Many fixes related to the ``cluster-merge`` tool
353 - Fixed a race condition in the lock monitor, which caused failures
354 during (at least) creation of many instances in parallel
355 - Improved output for gnt-job info
356 - Removed the quiet flag on some ssh calls which prevented debugging
358 - Improved the N+1 failure messages in cluster verify by actually
359 showing the memory values (needed and available)
360 - Increased lock attempt timeouts so that when executing long operations
361 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
362 too early and thus prevent the use of the 'fair' mechanism
363 - Changed instance query data (``gnt-instance info``) to not acquire
364 locks unless needed, thus allowing its use on locked instance if only
365 static information is asked for
366 - Improved behaviour with filesystems that do not support rename on an
368 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
369 kept locks on all nodes during the wipe, which is unneeded
370 - Fixed ``gnt-watcher`` handling of errors during hooks execution
371 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
372 10GiB) which caused the wipe to fail right at the end in some cases
373 - Fixed master IP activation when doing master failover with no-voting
374 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
375 the master node itself
376 - Fixed potential data-loss in under disk full conditions, where Ganeti
377 wouldn't check correctly the return code and would consider
378 partially-written files 'correct'
379 - Fixed bug related to multiple VGs and DRBD disk replacing
380 - Added new disk parameter ``metavg`` that allows placement of the meta
381 device for DRBD in a different volume group
382 - Fixed error handling in the node daemon when the system libc doesn't
383 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
384 - Fixed lock release during replace-disks, which kept cluster-wide locks
385 when doing disk replaces with an iallocator script
386 - Added check for missing bridges in cluster verify
387 - Handle EPIPE errors while writing to the terminal better, so that
388 piping the output to e.g. ``less`` doesn't cause a backtrace
389 - Fixed rare case where a ^C during Luxi calls could have been
390 interpreted as server errors, instead of simply terminating
391 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
393 - Added a few more parameters to the KVM hypervisor, allowing a second
394 CDROM, custom disk type for CDROMs and a floppy image
395 - Removed redundant message in instance rename when the name is given
397 - Added option to ``gnt-instance recreate-disks`` to allow creating the
398 disks on new nodes, allowing recreation when the original instance
399 nodes are completely gone
400 - Added option when converting disk templates to DRBD to skip waiting
401 for the resync, in order to make the instance available sooner
402 - Added two new variables to the OS scripts environment (containing the
404 - Made the root_path and optional parameter for the xen-pvm hypervisor,
405 to allow use of ``pvgrub`` as bootloader
406 - Changed the instance memory modifications to only check out-of-memory
407 conditions on memory increases, and turned the secondary node warnings
408 into errors (they can still be overridden via ``--force``)
409 - Fixed the handling of a corner case when the Python installation gets
410 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
411 to execute a command that doesn't exist
412 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
413 node of the instance returned failures during instance shutdown; this
414 adds the option ``--ignore-consistency`` to gnt-instance move
416 And as usual, various improvements to the error messages, documentation
423 *(Released Wed, 09 Mar 2011)*
425 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
426 the RAPI users file if run twice (even with ``--dry-run``).
428 The release fixes that bug (nothing else changed).
434 *(Released Mon, 07 Mar 2011)*
436 Final 2.4.0 release. Just a few small fixes:
438 - Fixed RAPI node evacuate
439 - Fixed the kvm-ifup script
440 - Fixed internal error handling for special job cases
441 - Updated man page to specify the escaping feature for options
447 *(Released Mon, 28 Feb 2011)*
449 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
450 that this feature wiped the disks of the wrong instance, leading to loss
455 - Fixed title of query field containing instance name
456 - Expanded the glossary in the documentation
457 - Fixed one unittest (internal issue)
463 *(Released Mon, 21 Feb 2011)*
465 A number of bug fixes plus just a couple functionality changes.
467 On the user-visible side, the ``gnt-* list`` command output has changed
468 with respect to "special" field states. The current rc1 style of display
469 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
470 the default output mode special fields states are displayed as follows:
472 - Offline resource: ``*``
473 - Unavailable/not applicable: ``-``
474 - Data missing (RPC failure): ``?``
475 - Unknown field: ``??``
477 Another user-visible change is the addition of ``--force-join`` to
482 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
483 - Fixed regression in RAPI/instance reinstall where all parameters were
484 required (instead of optional)
485 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
486 - Fixed iallocator usage (offline nodes were not considered offline)
487 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
488 - Fixed hypervisor and OS parameter validation with respect to
490 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
492 - Fixed ``tools/listrunner`` with respect to agent-based usage
498 *(Released Fri, 4 Feb 2011)*
500 Many changes and fixes since the beta1 release. While there were some
501 internal changes, the code has been mostly stabilised for the RC
504 Note: the dumb allocator was removed in this release, as it was not kept
505 up-to-date with the IAllocator protocol changes. It is recommended to
506 use the ``hail`` command from the ganeti-htools package.
508 Note: the 2.4 and up versions of Ganeti are not compatible with the
509 0.2.x branch of ganeti-htools. You need to upgrade to
510 ganeti-htools-0.3.0 (or later).
512 Regressions fixed from 2.3
513 ~~~~~~~~~~~~~~~~~~~~~~~~~~
515 - Fixed the ``gnt-cluster verify-disks`` command
516 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
518 - Fixed disk adoption breakage
519 - Fixed wrong headers in instance listing for field aliases
524 - Fixed corner case in KVM handling of NICs
525 - Fixed many cases of wrong handling of non-vm_capable nodes
526 - Fixed a bug where a missing instance symlink was not possible to
527 recreate with any ``gnt-*`` command (now ``gnt-instance
528 activate-disks`` does it)
529 - Fixed the volume group name as reported by ``gnt-cluster
531 - Increased timeouts for the import-export code, hopefully leading to
532 fewer aborts due network or instance timeouts
533 - Fixed bug in ``gnt-node list-storage``
534 - Fixed bug where not all daemons were started on cluster
535 initialisation, but only at the first watcher run
536 - Fixed many bugs in the OOB implementation
537 - Fixed watcher behaviour in presence of instances with offline
539 - Fixed instance list output for instances running on the wrong node
540 - a few fixes to the cluster-merge tool, but it still cannot merge
541 multi-node groups (currently it is not recommended to use this tool)
547 - Improved network configuration for the KVM hypervisor
548 - Added e1000 as a supported NIC for Xen-HVM
549 - Improved the lvmstrap tool to also be able to use partitions, as
550 opposed to full disks
551 - Improved speed of disk wiping (the cluster parameter
552 ``prealloc_wipe_disks``, so that it has a low impact on the total time
553 of instance creations
554 - Added documentation for the OS parameters
555 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
556 hypervisor is not responding
557 - Added display of blacklisted and hidden OS information in
559 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
560 NIC and node parameters, which might create problems with currently
561 invalid (but undetected) configuration files, but prevents validation
562 failures when unrelated parameters are modified
563 - Changed cluster initialisation to wait for the master daemon to become
565 - Expanded the RAPI interface:
567 - Added config redistribution resource
568 - Added activation/deactivation of instance disks
569 - Added export of console information
571 - Implemented log file reopening on SIGHUP, which allows using
572 logrotate(8) for the Ganeti log files
573 - Added a basic OOB helper script as an example
579 *(Released Fri, 14 Jan 2011)*
584 - Fixed timezone issues when formatting timestamps
585 - Added support for node groups, available via ``gnt-group`` and other
587 - Added out-of-band framework and management, see :doc:`design
588 document <design-oob>`
589 - Removed support for roman numbers from ``gnt-node list`` and
590 ``gnt-instance list``.
591 - Allowed modification of master network interface via ``gnt-cluster
592 modify --master-netdev``
593 - Accept offline secondaries while shutting down instance disks
594 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
595 - Added support for multiple LVM volume groups
596 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
598 - Added commands to list available fields:
600 - ``gnt-node list-fields``
601 - ``gnt-group list-fields``
602 - ``gnt-instance list-fields``
604 - Updated documentation and man pages
609 - Moved ``rapi_users`` file into separate directory, now named
610 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
612 - Added new tool for running commands on many machines,
613 ``tools/ganeti-listrunner``
614 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
615 improving the ``gnt-instance console`` output
616 - Allowed customisation of disk index separator at ``configure`` time
617 - Export node group allocation policy to :doc:`iallocator <iallocator>`
618 - Added support for non-partitioned md disks in ``lvmstrap``
619 - Added script to gracefully power off KVM instances
620 - Split ``utils`` module into smaller parts
621 - Changed query operations to return more detailed information, e.g.
622 whether an information is unavailable due to an offline node. To use
623 this new functionality, the LUXI call ``Query`` must be used. Field
624 information is now stored by the master daemon and can be retrieved
625 using ``QueryFields``. Instances, nodes and groups can also be queried
626 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
627 exposed via RAPI). The following commands make use of this
628 infrastructure change:
631 - ``gnt-group list-fields``
633 - ``gnt-node list-fields``
634 - ``gnt-instance list``
635 - ``gnt-instance list-fields``
636 - ``gnt-debug locks``
641 - New RAPI resources (see :doc:`rapi`):
645 - ``/2/groups/[group_name]``
646 - ``/2/groups/[group_name]/assign-nodes``
647 - ``/2/groups/[group_name]/modify``
648 - ``/2/groups/[group_name]/rename``
649 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
653 - Implemented ``no_install`` for instance creation
654 - Implemented OS parameters for instance reinstallation, allowing
655 use of special settings on reinstallation (e.g. for preserving data)
660 - Added IPv6 support in import/export
661 - Pause DRBD synchronization while wiping disks on instance creation
662 - Updated unittests and QA scripts
663 - Improved network parameters passed to KVM
664 - Converted man pages from docbook to reStructuredText
670 *(Released Mon, 20 Dec 2010)*
672 Released version 2.3.1~rc1 without any changes.
678 *(Released Wed, 1 Dec 2010)*
680 - impexpd: Disable OpenSSL compression in socat if possible (backport
681 from master, commit e90739d625b, see :doc:`installation guide
682 <install-quick>` for details)
683 - Changed unittest coverage report to exclude test scripts
684 - Added script to check version format
690 *(Released Wed, 1 Dec 2010)*
692 Released version 2.3.0~rc1 without any changes.
698 *(Released Fri, 19 Nov 2010)*
700 A number of bugfixes and documentation updates:
702 - Update ganeti-os-interface documentation
703 - Fixed a bug related to duplicate MACs or similar items which should be
705 - Fix breakage in OS state modify
706 - Reinstall instance: disallow offline secondaries (fixes bug related to
707 OS changing but reinstall failing)
708 - plus all the other fixes between 2.2.1 and 2.2.2
714 *(Released Tue, 2 Nov 2010)*
716 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
717 - Fixed master failover race with watcher
718 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
720 - Accept previously stopped instance for export with instance removal
721 - Simplify and extend the environment variables for instance OS scripts
722 - Added new node flags, ``master_capable`` and ``vm_capable``
723 - Added optional instance disk wiping prior during allocation. This is a
724 cluster-wide option and can be set/modified using
725 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
726 - Added IPv6 support, see :doc:`design document <design-2.3>` and
728 - Added a new watcher option (``--ignore-pause``)
729 - Added option to ignore offline node on instance start/stop
730 (``--ignore-offline``)
731 - Allow overriding OS parameters with ``gnt-instance reinstall``
732 - Added ability to change node's secondary IP address using ``gnt-node
734 - Implemented privilege separation for all daemons except
735 ``ganeti-noded``, see ``configure`` options
736 - Complain if an instance's disk is marked faulty in ``gnt-cluster
738 - Implemented job priorities (see ``ganeti(7)`` manpage)
739 - Ignore failures while shutting down instances during failover from
741 - Exit daemon's bootstrap process only once daemon is ready
742 - Export more information via ``LUInstanceQuery``/remote API
743 - Improved documentation, QA and unittests
744 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
745 restart if the file was created or changed
746 - Added LUXI protocol version sent with each request and response,
747 allowing detection of server/client mismatches
748 - Moved the Python scripts among gnt-* and ganeti-* into modules
749 - Moved all code related to setting up SSH to an external script,
751 - Infrastructure changes for node group support in future versions
757 *(Released Fri, 19 Nov 2010)*
759 A few small bugs fixed, and some improvements to the build system:
761 - Fix documentation regarding conversion to drbd
762 - Fix validation of parameters in cluster modify (``gnt-cluster modify
764 - Fix error handling in node modify with multiple changes
765 - Allow remote imports without checked names
771 *(Released Tue, 19 Oct 2010)*
773 - Disable SSL session ID cache in RPC client
779 *(Released Thu, 14 Oct 2010)*
781 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
782 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
784 - Fix problems with interaction between Curl and Python's HTTP server,
785 resulting in increased speed in many RPC calls
786 - Improve our release script to prevent breakage with older aclocal and
793 *(Released Thu, 7 Oct 2010)*
795 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
796 value retrieved from master
797 - Added support for blacklisted or hidden OS definitions
798 - Added simple lock monitor (accessible via (``gnt-debug locks``)
799 - Added support for -mem-path in KVM hypervisor abstraction layer
800 - Allow overriding instance parameters in tool for inter-cluster
801 instance moves (``tools/move-instance``)
802 - Improved opcode summaries (e.g. in ``gnt-job list``)
803 - Improve consistency of OS listing by sorting it
804 - Documentation updates
810 *(Released Fri, 8 Oct 2010)*
812 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
818 *(Released Mon, 4 Oct 2010)*
820 - Fixed regression in ``gnt-instance rename``
826 *(Released Wed, 22 Sep 2010)*
828 - Fixed OS_VARIANT variable for OS scripts
829 - Fixed cluster tag operations via RAPI
830 - Made ``setup-ssh`` exit with non-zero code if an error occurred
831 - Disabled RAPI CA checks in watcher
837 *(Released Mon, 23 Aug 2010)*
839 - Support DRBD versions of the format "a.b.c.d"
841 - Re-introduce support for usage from multiple threads in RAPI client
842 - Instance renames and modify via RAPI
843 - Work around race condition between processing and archival in job
845 - Mark opcodes following failed one as failed, too
846 - Job field ``lock_status`` was removed due to difficulties making it
847 work with the changed job queue in Ganeti 2.2; a better way to monitor
848 locks is expected for a later 2.2.x release
849 - Fixed dry-run behaviour with many commands
850 - Support ``ssh-agent`` again when adding nodes
851 - Many additional bugfixes
857 *(Released Fri, 30 Jul 2010)*
859 Important change: the internal RPC mechanism between Ganeti nodes has
860 changed from using a home-grown http library (based on the Python base
861 libraries) to use the PycURL library. This requires that PycURL is
862 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
863 against GnuTLS by default. cURL's support for GnuTLS had known issues
864 before cURL 7.21.0 and we recommend using the latest cURL release or
865 linking against OpenSSL. Most other distributions already link PycURL
866 and cURL against OpenSSL. The command::
868 python -c 'import pycurl; print pycurl.version'
870 can be used to determine the libraries PycURL and cURL are linked
873 Other significant changes:
875 - Rewrote much of the internals of the job queue, in order to achieve
876 better parallelism; this decouples job query operations from the job
877 processing, and it should allow much nicer behaviour of the master
878 daemon under load, and it also has uncovered some long-standing bugs
879 related to the job serialisation (now fixed)
880 - Added a default iallocator setting to the cluster parameters,
881 eliminating the need to always pass nodes or an iallocator for
882 operations that require selection of new node(s)
883 - Added experimental support for the LXC virtualization method
884 - Added support for OS parameters, which allows the installation of
885 instances to pass parameter to OS scripts in order to customise the
887 - Added a hypervisor parameter controlling the migration type (live or
888 non-live), since hypervisors have various levels of reliability; this
889 has renamed the 'live' parameter to 'mode'
890 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
891 logical volumes, meaning that cluster verify will ignore them and not
892 flag their presence as errors
893 - The watcher will now reset the error count for failed instances after
894 8 hours, thus allowing self-healing if the problem that caused the
895 instances to be down/fail to start has cleared in the meantime
896 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
897 check for, and warn, if the drbd module parameter ``usermode_helper``
898 is not consistent with the cluster-wide setting; this is needed to
899 make diagnose easier of failed drbd creations
900 - Started adding base IPv6 support, but this is not yet
901 enabled/available for use
902 - Rename operations (cluster, instance) will now return the new name,
903 which is especially useful if a short name was passed in
904 - Added support for instance migration in RAPI
905 - Added a tool to pre-configure nodes for the SSH setup, before joining
906 them to the cluster; this will allow in the future a simplified model
907 for node joining (but not yet fully enabled in 2.2); this needs the
908 paramiko python library
909 - Fixed handling of name-resolving errors
910 - Fixed consistency of job results on the error path
911 - Fixed master-failover race condition when executed multiple times in
913 - Fixed many bugs related to the job queue (mostly introduced during the
914 2.2 development cycle, so not all are impacting 2.1)
915 - Fixed instance migration with missing disk symlinks
916 - Fixed handling of unknown jobs in ``gnt-job archive``
917 - And many other small fixes/improvements
921 - Enhanced both the unittest and the QA coverage
922 - Switched the opcode validation to a generic model, and extended the
923 validation to all opcode parameters
924 - Changed more parts of the code that write shell scripts to use the
926 - Switched the master daemon to use the asyncore library for the Luxi
933 *(Released Thu, 17 Jun 2010)*
935 - Added tool (``move-instance``) and infrastructure to move instances
936 between separate clusters (see :doc:`separate documentation
937 <move-instance>` and :doc:`design document <design-2.2>`)
938 - Added per-request RPC timeout
939 - RAPI now requires a Content-Type header for requests with a body (e.g.
940 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
941 :rfc:`2616` (HTTP/1.1), section 7.2.1)
942 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
944 - Implemented initial support for running Ganeti daemons as separate
945 users, see configure-time flags ``--with-user-prefix`` and
946 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
948 - Instances can be removed after export (``gnt-backup export
950 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
951 key (instead of 1024 bit)
952 - Added new cluster configuration file for cluster domain secret
953 - Import/export now use SSL instead of SSH
954 - Added support for showing estimated time when exporting an instance,
955 see the ``ganeti-os-interface(7)`` manpage and look for
962 *(Released Tue, 16 Nov 2010)*
964 Some more bugfixes. Unless critical bugs occur, this will be the last
967 - Fix case of MAC special-values
968 - Fix mac checker regex
969 - backend: Fix typo causing "out of range" error
970 - Add missing --units in gnt-instance list man page
976 *(Released Tue, 24 Aug 2010)*
979 - Don't ignore secondary node silently on non-mirrored disk templates
981 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
982 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
983 - Properly document the usb_mouse parameter
984 - Fix path in ganeti-rapi(8) (issue 116)
985 - Adjust error message when the ganeti user's .ssh directory is
987 - Add same-node-check when changing the disk template to drbd
993 *(Released Fri, 16 Jul 2010)*
996 - Add an option to only select some reboot types during qa/burnin.
997 (on some hypervisors consequent reboots are not supported)
998 - Fix infrequent race condition in master failover. Sometimes the old
999 master ip address would be still detected as up for a short time
1000 after it was removed, causing failover to fail.
1001 - Decrease mlockall warnings when the ctypes module is missing. On
1002 Python 2.4 we support running even if no ctypes module is installed,
1003 but we were too verbose about this issue.
1004 - Fix building on old distributions, on which man doesn't have a
1006 - Fix RAPI not to ignore the MAC address on instance creation
1007 - Implement the old instance creation format in the RAPI client.
1013 *(Released Thu, 01 Jul 2010)*
1015 A small bugfix release:
1016 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1017 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1018 a non-existing option
1019 - Fix instance create: the --force-variant option was ignored
1020 - Improve pylint 0.21 compatibility and warnings with Python 2.6
1021 - Fix modify node storage with non-FQDN arguments
1022 - Fix RAPI client to authenticate under Python 2.6 when used
1023 for more than 5 requests needing authentication
1024 - Fix gnt-instance modify -t (storage) giving a wrong error message
1025 when converting a non-shutdown drbd instance to plain
1031 *(Released Fri, 18 Jun 2010)*
1033 A small bugfix release:
1035 - Fix live migration of KVM instances started with older Ganeti
1036 versions which had fewer hypervisor parameters
1037 - Fix gnt-instance grow-disk on down instances
1038 - Fix an error-reporting bug during instance migration
1039 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1040 silently ignoring broken ones
1041 - Fix an RPC error reporting bug affecting, for example, RAPI client
1043 - Fix bug triggered by different API version os-es on different nodes
1044 - Fix a bug in instance startup with custom hvparams: OS level
1045 parameters would fail to be applied.
1046 - Fix the RAPI client under Python 2.6 (but more work is needed to
1047 make it work completely well with OpenSSL)
1048 - Fix handling of errors when resolving names from DNS
1054 *(Released Thu, 3 Jun 2010)*
1056 A medium sized development cycle. Some new features, and some
1057 fixes/small improvements/cleanups.
1059 Significant features
1060 ~~~~~~~~~~~~~~~~~~~~
1062 The node deamon now tries to mlock itself into memory, unless the
1063 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1064 its logs, and falls back to console logging. This allows emergency
1065 features such as ``gnt-node powercycle`` to work even in the event of a
1066 broken node disk (tested offlining the disk hosting the node's
1067 filesystem and dropping its memory caches; don't try this at home)
1069 KVM: add vhost-net acceleration support. It can be tested with a new
1070 enough version of the kernel and of qemu-kvm.
1072 KVM: Add instance chrooting feature. If you use privilege dropping for
1073 your VMs you can also now force them to chroot to an empty directory,
1074 before starting the emulated guest.
1076 KVM: Add maximum migration bandwith and maximum downtime tweaking
1077 support (requires a new-enough version of qemu-kvm).
1079 Cluster verify will now warn if the master node doesn't have the master
1080 ip configured on it.
1082 Add a new (incompatible) instance creation request format to RAPI which
1083 supports all parameters (previously only a subset was supported, and it
1084 wasn't possible to extend the old format to accomodate all the new
1085 features. The old format is still supported, and a client can check for
1086 this feature, before using it, by checking for its presence in the
1087 ``features`` RAPI resource.
1089 Now with ancient latin support. Try it passing the ``--roman`` option to
1090 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1091 (requires the python-roman module to be installed, in order to work).
1096 As usual many internal code refactorings, documentation updates, and
1099 - Lots of improvements and cleanups to the experimental Remote API
1100 (RAPI) client library.
1101 - A new unit test suite for the core daemon libraries.
1102 - A fix to creating missing directories makes sure the umask is not
1103 applied anymore. This enforces the same directory permissions
1105 - Better handling terminating daemons with ctrl+c (used when running
1106 them in debugging mode).
1107 - Fix a race condition in live migrating a KVM instance, when stat()
1108 on the old proc status file returned EINVAL, which is an unexpected
1110 - Fixed manpage checking with newer man and utf-8 charachters. But now
1111 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1117 *(Released Fri, 7 May 2010)*
1119 Fix a bug which prevented untagged KVM instances from starting.
1125 *(Released Fri, 7 May 2010)*
1127 Another release with a long development cycle, during which many
1128 different features were added.
1130 Significant features
1131 ~~~~~~~~~~~~~~~~~~~~
1133 The KVM hypervisor now can run the individual instances as non-root, to
1134 reduce the impact of a VM being hijacked due to bugs in the
1135 hypervisor. It is possible to run all instances as a single (non-root)
1136 user, to manually specify a user for each instance, or to dynamically
1137 allocate a user out of a cluster-wide pool to each instance, with the
1138 guarantee that no two instances will run under the same user ID on any
1141 An experimental RAPI client library, that can be used standalone
1142 (without the other Ganeti libraries), is provided in the source tree as
1143 ``lib/rapi/client.py``. Note this client might change its interface in
1144 the future, as we iterate on its capabilities.
1146 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1147 replace the cluster's certificates and crypto keys. This might help in
1148 case they have been compromised, or have simply expired.
1150 A new disk option for instance creation has been added that allows one
1151 to "adopt" currently existing logical volumes, with data
1152 preservation. This should allow easier migration to Ganeti from
1153 unmanaged (or managed via other software) instances.
1155 Another disk improvement is the possibility to convert between redundant
1156 (DRBD) and plain (LVM) disk configuration for an instance. This should
1157 allow better scalability (starting with one node and growing the
1158 cluster, or shrinking a two-node cluster to one node).
1160 A new feature that could help with automated node failovers has been
1161 implemented: if a node sees itself as offline (by querying the master
1162 candidates), it will try to shutdown (hard) all instances and any active
1163 DRBD devices. This reduces the risk of duplicate instances if an
1164 external script automatically failovers the instances on such nodes. To
1165 enable this, the cluster parameter ``maintain_node_health`` should be
1166 enabled; in the future this option (per the name) will enable other
1167 automatic maintenance features.
1169 Instance export/import now will reuse the original instance
1170 specifications for all parameters; that means exporting an instance,
1171 deleting it and the importing it back should give an almost identical
1172 instance. Note that the default import behaviour has changed from
1173 before, where it created only one NIC; now it recreates the original
1176 Cluster verify has added a few new checks: SSL certificates validity,
1177 /etc/hosts consistency across the cluster, etc.
1182 As usual, many internal changes were done, documentation fixes,
1185 - Fixed cluster initialization with disabled cluster storage (regression
1186 introduced in 2.1.1)
1187 - File-based storage supports growing the disks
1188 - Fixed behaviour of node role changes
1189 - Fixed cluster verify for some corner cases, plus a general rewrite of
1190 cluster verify to allow future extension with more checks
1191 - Fixed log spamming by watcher and node daemon (regression introduced
1193 - Fixed possible validation issues when changing the list of enabled
1195 - Fixed cleanup of /etc/hosts during node removal
1196 - Fixed RAPI response for invalid methods
1197 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1198 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1199 from ide disks, etc.)
1200 - Allow OS changes without re-installation (to record a changed OS
1201 outside of Ganeti, or to allow OS renames)
1202 - Allow instance creation without OS installation (useful for example if
1203 the OS will be installed manually, or restored from a backup not in
1205 - Implemented option to make cluster ``copyfile`` use the replication
1207 - Added list of enabled hypervisors to ssconf (possibly useful for
1209 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1211 - A partial form of node re-IP is possible via node readd, which now
1212 allows changed node primary IP
1213 - Command line utilities now show an informational message if the job is
1215 - The logs of the master daemon now show the PID/UID/GID of the
1222 *(Released Fri, 12 Mar 2010)*
1224 During the 2.1.0 long release candidate cycle, a lot of improvements and
1225 changes have accumulated with were released later as 2.1.1.
1230 The node evacuate command (``gnt-node evacuate``) was significantly
1231 rewritten, and as such the IAllocator protocol was changed - a new
1232 request type has been added. This unfortunate change during a stable
1233 series is designed to improve performance of node evacuations; on
1234 clusters with more than about five nodes and which are well-balanced,
1235 evacuation should proceed in parallel for all instances of the node
1236 being evacuated. As such, any existing IAllocator scripts need to be
1237 updated, otherwise the above command will fail due to the unknown
1238 request. The provided "dumb" allocator has not been updated; but the
1239 ganeti-htools package supports the new protocol since version 0.2.4.
1241 Another important change is increased validation of node and instance
1242 names. This might create problems in special cases, if invalid host
1243 names are being used.
1245 Also, a new layer of hypervisor parameters has been added, that sits at
1246 OS level between the cluster defaults and the instance ones. This allows
1247 customisation of virtualization parameters depending on the installed
1248 OS. For example instances with OS 'X' may have a different KVM kernel
1249 (or any other parameter) than the cluster defaults. This is intended to
1250 help managing a multiple OSes on the same cluster, without manual
1251 modification of each instance's parameters.
1253 A tool for merging clusters, ``cluster-merge``, has been added in the
1254 tools sub-directory.
1259 - Improved the int/float conversions that should make the code more
1260 robust in face of errors from the node daemons
1261 - Fixed the remove node code in case of internal configuration errors
1262 - Fixed the node daemon behaviour in face of inconsistent queue
1263 directory (e.g. read-only file-system where we can't open the files
1265 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1266 now it either aborts cleanly or, if given the new "auto_promote"
1267 parameter, will automatically promote other nodes as needed
1268 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1269 completely prevent Ganeti from working
1270 - Fixed bug for instance export when not all disks were successfully
1272 - Fixed behaviour of node add when the new node is slow in starting up
1274 - Fixed handling of signals in the LUXI client, which should improve
1275 behaviour of command-line scripts
1276 - Added checks for invalid node/instance names in the configuration (now
1277 flagged during cluster verify)
1278 - Fixed watcher behaviour for disk activation errors
1279 - Fixed two potentially endless loops in http library, which led to the
1280 RAPI daemon hanging and consuming 100% CPU in some cases
1281 - Fixed bug in RAPI daemon related to hashed passwords
1282 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1284 - Enhanced compatibility with non-Debian OSes, but not using absolute
1285 path in some commands and allowing customisation of the ssh
1286 configuration directory
1287 - Fixed possible future issue with new Python versions by abiding to the
1288 proper use of ``__slots__`` attribute on classes
1289 - Added checks that should prevent directory traversal attacks
1290 - Many documentation fixes based on feedback from users
1295 - Added an "early_release" more for instance replace disks and node
1296 evacuate, where we release locks earlier and thus allow higher
1297 parallelism within the cluster
1298 - Added watcher hooks, intended to allow the watcher to restart other
1299 daemons (e.g. from the ganeti-nbma project), but they can be used of
1300 course for any other purpose
1301 - Added a compile-time disable for DRBD barriers, to increase
1302 performance if the administrator trusts the power supply or the
1303 storage system to not lose writes
1304 - Added the option of using syslog for logging instead of, or in
1305 addition to, Ganeti's own log files
1306 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1307 can indeed boot from a paravirtual NIC
1308 - Added a generic debug level for many operations; while this is not
1309 used widely yet, it allows one to pass the debug value all the way to
1311 - Enhanced the hooks environment for instance moves (failovers,
1312 migrations) where the primary/secondary nodes changed during the
1313 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1314 - Enhanced data validations for many user-supplied values; one important
1315 item is the restrictions imposed on instance and node names, which
1316 might reject some (invalid) host names
1317 - Add a configure-time option to disable file-based storage, if it's not
1318 needed; this allows greater security separation between the master
1319 node and the other nodes from the point of view of the inter-node RPC
1321 - Added user notification in interactive tools if job is waiting in the
1322 job queue or trying to acquire locks
1323 - Added log messages when a job is waiting for locks
1324 - Added filtering by node tags in instance operations which admit
1325 multiple instances (start, stop, reboot, reinstall)
1326 - Added a new tool for cluster mergers, ``cluster-merge``
1327 - Parameters from command line which are of the form ``a=b,c=d`` can now
1328 use backslash escapes to pass in values which contain commas,
1329 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1331 - For KVM, the instance name is the first parameter passed to KVM, so
1332 that it's more visible in the process list
1338 *(Released Tue, 2 Mar 2010)*
1340 Ganeti 2.1 brings many improvements with it. Major changes:
1342 - Added infrastructure to ease automated disk repairs
1343 - Added new daemon to export configuration data in a cheaper way than
1344 using the remote API
1345 - Instance NICs can now be routed instead of being associated with a
1347 - Improved job locking logic to reduce impact of jobs acquiring multiple
1348 locks waiting for other long-running jobs
1350 In-depth implementation details can be found in the Ganeti 2.1 design
1356 - Added chroot hypervisor
1357 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1359 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1360 ``bootloader_path`` and ``bootloader_args``)
1361 - Added the ``use_localtime`` option for the xen-hvm and kvm
1362 hypervisors, and the default value for this has changed to false (in
1363 2.0 xen-hvm always enabled it)
1364 - Added luxi call to submit multiple jobs in one go
1365 - Added cluster initialization option to not modify ``/etc/hosts``
1367 - Added network interface parameters
1368 - Added dry run mode to some LUs
1369 - Added RAPI resources:
1371 - ``/2/instances/[instance_name]/info``
1372 - ``/2/instances/[instance_name]/replace-disks``
1373 - ``/2/nodes/[node_name]/evacuate``
1374 - ``/2/nodes/[node_name]/migrate``
1375 - ``/2/nodes/[node_name]/role``
1376 - ``/2/nodes/[node_name]/storage``
1377 - ``/2/nodes/[node_name]/storage/modify``
1378 - ``/2/nodes/[node_name]/storage/repair``
1380 - Added OpCodes to evacuate or migrate all instances on a node
1381 - Added new command to list storage elements on nodes (``gnt-node
1382 list-storage``) and modify them (``gnt-node modify-storage``)
1383 - Added new ssconf files with master candidate IP address
1384 (``ssconf_master_candidates_ips``), node primary IP address
1385 (``ssconf_node_primary_ips``) and node secondary IP address
1386 (``ssconf_node_secondary_ips``)
1387 - Added ``ganeti-confd`` and a client library to query the Ganeti
1388 configuration via UDP
1389 - Added ability to run hooks after cluster initialization and before
1391 - Added automatic mode for disk replace (``gnt-instance replace-disks
1393 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1394 after catastrophic data-loss
1395 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1397 - Added ``gnt-instance move`` command to move instances
1398 - Added ``gnt-cluster watcher`` command to control watcher
1399 - Added ``gnt-node powercycle`` command to powercycle nodes
1400 - Added new job status field ``lock_status``
1401 - Added parseable error codes to cluster verification (``gnt-cluster
1402 verify --error-codes``) and made output less verbose (use
1403 ``--verbose`` to restore previous behaviour)
1404 - Added UUIDs to the main config entities (cluster, nodes, instances)
1405 - Added support for OS variants
1406 - Added support for hashed passwords in the Ganeti remote API users file
1408 - Added option to specify maximum timeout on instance shutdown
1409 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1410 - Added new helper script to start and stop Ganeti daemons
1411 (``daemon-util``), with the intent to reduce the work necessary to
1412 adjust Ganeti for non-Debian distributions and to start/stop daemons
1414 - Added more unittests
1415 - Fixed critical bug in ganeti-masterd startup
1416 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1417 now customisable at the cluster level for both the KVM and Xen
1418 hypervisors using the new ``migration_port`` parameter
1419 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1420 reinstalling an instance
1421 - Allowed ``@`` in tag names
1422 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1423 - Many documentation updates
1424 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1425 - ``gnt-instance reinstall`` can now reinstall multiple instances
1426 - Updated many command line parameters
1427 - Introduced new OS API version 15
1428 - No longer support a default hypervisor
1429 - Treat virtual LVs as inexistent
1430 - Improved job locking logic to reduce lock contention
1431 - Match instance and node names case insensitively
1432 - Reimplemented bash completion script to be more complete
1439 *(Released Thu, 4 Feb 2010)*
1441 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1442 - Fix a string formatting bug
1443 - Improve safety of the code in some error paths
1444 - Improve data validation in the master of values returned from nodes
1450 *(Released Thu, 17 Dec 2009)*
1452 - Fix security issue due to missing validation of iallocator names; this
1453 allows local and remote execution of arbitrary executables
1454 - Fix failure of gnt-node list during instance removal
1455 - Ship the RAPI documentation in the archive
1461 *(Released Wed, 30 Sep 2009)*
1463 - Fixed many wrong messages
1464 - Fixed a few bugs related to the locking library
1465 - Fixed MAC checking at instance creation time
1466 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1467 - Fixed a few issues related to signal handling in both daemons and
1469 - Fixed the example startup script provided
1470 - Fixed insserv dependencies in the example startup script (patch from
1472 - Fixed handling of drained nodes in the iallocator framework
1473 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1475 - Fixed error related to invalid job IDs in job polling
1476 - Fixed job/opcode persistence on unclean master shutdown
1477 - Fixed handling of partial job processing after unclean master
1479 - Fixed error reporting from LUs, previously all errors were converted
1480 into execution errors
1481 - Fixed error reporting from burnin
1482 - Decreased significantly the memory usage of the job queue
1483 - Optimised slightly multi-job submission
1484 - Optimised slightly opcode loading
1485 - Backported the multi-job submit framework from the development
1486 branch; multi-instance start and stop should be faster
1487 - Added script to clean archived jobs after 21 days; this will reduce
1488 the size of the queue directory
1489 - Added some extra checks in disk size tracking
1490 - Added an example ethers hook script
1491 - Added a cluster parameter that prevents Ganeti from modifying of
1493 - Added more node information to RAPI responses
1494 - Added a ``gnt-job watch`` command that allows following the ouput of a
1496 - Added a bind-address option to ganeti-rapi
1497 - Added more checks to the configuration verify
1498 - Enhanced the burnin script such that some operations can be retried
1500 - Converted instance reinstall to multi-instance model
1506 *(Released Fri, 7 Aug 2009)*
1508 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1509 to allow using the pre-2.0.2 behaviour in activation, if any existing
1510 instances have mismatched disk sizes in the configuration
1511 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1512 any configuration mismatches for disk sizes
1513 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1514 failover to work on two-node clusters
1515 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1517 - Fixed detection of OS script errors in ``gnt-backup export``
1518 - Fixed exit code of ``gnt-backup export``
1524 *(Released Fri, 17 Jul 2009)*
1526 - Added experimental support for stripped logical volumes; this should
1527 enhance performance but comes with a higher complexity in the block
1528 device handling; stripping is only enabled when passing
1529 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1530 affected even in the non-stripped mode
1531 - Improved resiliency against transient failures at the end of DRBD
1532 resyncs, and in general of DRBD resync checks
1533 - Fixed a couple of issues with exports and snapshot errors
1534 - Fixed a couple of issues in instance listing
1535 - Added display of the disk size in ``gnt-instance info``
1536 - Fixed checking for valid OSes in instance creation
1537 - Fixed handling of the "vcpus" parameter in instance listing and in
1538 general of invalid parameters
1539 - Fixed http server library, and thus RAPI, to handle invalid
1540 username/password combinations correctly; this means that now they
1541 report unauthorized for queries too, not only for modifications,
1542 allowing earlier detect of configuration problems
1543 - Added a new "role" node list field, equivalent to the master/master
1544 candidate/drained/offline flags combinations
1545 - Fixed cluster modify and changes of candidate pool size
1546 - Fixed cluster verify error messages for wrong files on regular nodes
1547 - Fixed a couple of issues with node demotion from master candidate role
1548 - Fixed node readd issues
1549 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1550 - Added a new ``--no-voting`` option for masterfailover to fix failover
1551 on two-nodes clusters when the former master node is unreachable
1552 - Added instance reinstall over RAPI
1558 *(Released Tue, 16 Jun 2009)*
1560 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1561 allow re-adding the start in single-user option (regression from 1.2)
1562 - the watcher writes the instance status to a file, to allow monitoring
1563 to report the instance status (from the master) based on cached
1564 results of the watcher's queries; while this can get stale if the
1565 watcher is being locked due to other work on the cluster, this is
1566 still an improvement
1567 - the watcher now also restarts the node daemon and the rapi daemon if
1569 - fixed the watcher to handle full and drained queue cases
1570 - hooks export more instance data in the environment, which helps if
1571 hook scripts need to take action based on the instance's properties
1572 (no longer need to query back into ganeti)
1573 - instance failovers when the instance is stopped do not check for free
1574 RAM, so that failing over a stopped instance is possible in low memory
1576 - rapi uses queries for tags instead of jobs (for less job traffic), and
1577 for cluster tags it won't talk to masterd at all but read them from
1579 - a couple of error handling fixes in RAPI
1580 - drbd handling: improved the error handling of inconsistent disks after
1581 resync to reduce the frequency of "there are some degraded disks for
1582 this instance" messages
1583 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1584 error handling path called a wrong function name)
1590 *(Released Wed, 27 May 2009)*
1592 - no changes from rc5
1598 *(Released Wed, 20 May 2009)*
1600 - fix a couple of bugs (validation, argument checks)
1601 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1602 - some small improvements to RAPI and IAllocator
1603 - make watcher automatically start the master daemon if down
1609 *(Released Mon, 27 Apr 2009)*
1611 - change the OS list to not require locks; this helps with big clusters
1612 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1613 volume group is broken
1614 - ``gnt-instance info``, without any arguments, doesn't run for all
1615 instances anymore; either pass ``--all`` or pass the desired
1616 instances; this helps against mistakes on big clusters where listing
1617 the information for all instances takes a long time
1618 - miscellaneous doc and man pages fixes
1624 *(Released Wed, 8 Apr 2009)*
1626 - Change the internal locking model of some ``gnt-node`` commands, in
1627 order to reduce contention (and blocking of master daemon) when
1628 batching many creation/reinstall jobs
1629 - Fixes to Xen soft reboot
1630 - No longer build documentation at build time, instead distribute it in
1631 the archive, in order to reduce the need for the whole docbook/rst
1638 *(Released Fri, 27 Mar 2009)*
1640 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1641 - Fix watcher startup sequence, improves the behaviour of busy clusters
1642 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1643 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1644 bind address and other places
1645 - Some documentation fixes and updates
1651 *(Released Mon, 2 Mar 2009)*
1653 - More documentation updates, now all docs should be more-or-less
1655 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1657 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1659 - Fix an issue related to $libdir/run/ganeti and cluster creation
1665 *(Released Thu, 19 Feb 2009)*
1667 - Xen PVM and KVM have switched the default value for the instance root
1668 disk to the first partition on the first drive, instead of the whole
1669 drive; this means that the OS installation scripts must be changed
1671 - Man pages have been updated
1672 - RAPI has been switched by default to HTTPS, and the exported functions
1673 should all work correctly
1674 - RAPI v1 has been removed
1675 - Many improvements to the KVM hypervisor
1676 - Block device errors are now better reported
1677 - Many other bugfixes and small improvements
1683 *(Released Mon, 26 Jan 2009)*
1685 - Version 2 is a general rewrite of the code and therefore the
1686 differences are too many to list, see the design document for 2.0 in
1687 the ``doc/`` subdirectory for more details
1688 - In this beta version there is not yet a migration path from 1.2 (there
1689 will be one in the final 2.0 release)
1690 - A few significant changes are:
1692 - all commands are executed by a daemon (``ganeti-masterd``) and the
1693 various ``gnt-*`` commands are just front-ends to it
1694 - all the commands are entered into, and executed from a job queue,
1695 see the ``gnt-job(8)`` manpage
1696 - the RAPI daemon supports read-write operations, secured by basic
1697 HTTP authentication on top of HTTPS
1698 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1699 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1700 to migrate to DRBD 8 first while still running Ganeti 1.2)
1701 - DRBD devices are using statically allocated minor numbers, which
1702 will be assigned to existing instances during the migration process
1703 - there is support for both Xen PVM and Xen HVM instances running on
1705 - KVM virtualization is supported too
1706 - file-based storage has been implemented, which means that it is
1707 possible to run the cluster without LVM and DRBD storage, for
1708 example using a shared filesystem exported from shared storage (and
1709 still have live migration)
1715 *(Released Tue, 13 Jan 2009)*
1717 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1718 - Reuse the old instance mac address by default on instance import, if
1719 the instance name is the same.
1720 - Handle situations in which the node info rpc returns incomplete
1722 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1723 - Improved version of batcher:
1725 - state file support
1726 - instance mac address support
1727 - support for HVM clusters/instances
1729 - Add an option to show the number of cpu sockets and nodes in
1731 - Support OSes that handle more than one version of the OS api (but do
1732 not change the current API in any other way)
1733 - Fix ``gnt-node migrate``
1734 - ``gnt-debug`` man page
1735 - Fixes various more typos and small issues
1736 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1742 *(Released Wed, 24 Sep 2008)*
1744 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1745 type of disk exported to fully virtualized instances.
1746 - provide access to the serial console of HVM instances
1747 - instance auto_balance flag, set by default. If turned off it will
1748 avoid warnings on cluster verify if there is not enough memory to fail
1749 over an instance. in the future it will prevent automatically failing
1750 it over when we will support that.
1751 - batcher tool for instance creation, see ``tools/README.batcher``
1752 - ``gnt-instance reinstall --select-os`` to interactively select a new
1753 operating system when reinstalling an instance.
1754 - when changing the memory amount on instance modify a check has been
1755 added that the instance will be able to start. also warnings are
1756 emitted if the instance will not be able to fail over, if auto_balance
1758 - documentation fixes
1759 - sync fields between ``gnt-instance list/modify/add/import``
1760 - fix a race condition in drbd when the sync speed was set after giving
1761 the device a remote peer.
1767 *(Released Tue, 22 Jul 2008)*
1769 - note: the allowed size and number of tags per object were reduced
1770 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1771 - fixed twisted 8.x compatibility
1772 - fixed ``gnt-instance replace-disks`` with iallocator
1773 - add TCP keepalives on twisted connections to detect restarted nodes
1774 - disk increase support, see ``gnt-instance grow-disk``
1775 - implement bulk node/instance query for RAPI
1776 - add tags in node/instance listing (optional)
1777 - experimental migration (and live migration) support, read the man page
1778 for ``gnt-instance migrate``
1779 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1780 has some small improvements in handling its state file
1786 *(Released Fri, 13 Jun 2008)*
1788 - Experimental readonly, REST-based remote API implementation;
1789 automatically started on master node, TCP port 5080, if enabled by
1790 ``--enable-rapi`` parameter to configure script.
1791 - Instance allocator support. Add and import instance accept a
1792 ``--iallocator`` parameter, and call that instance allocator to decide
1793 which node to use for the instance. The iallocator document describes
1794 what's expected from an allocator script.
1795 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1796 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1797 node is lost there is still enough memory to fail over the instances
1799 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1800 ``gnt-cluster verify``, to check for site-specific compliance. All the
1801 hooks will run, and their output, if any, will be displayed. Any
1802 failing hook will make the verification return an error value.
1803 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1804 primary and secondary interfaces
1805 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1806 that is still declared as part of the cluster and has failed.
1807 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1808 specifying output fields, that just adds the chosen fields to the
1810 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1811 export from the filesystem.
1812 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1813 have been added. Using them you can enable/disable acpi and pae
1814 support, and specify a path for a cd image to be exported to the
1815 instance. These parameters as the name suggest only work on HVM
1817 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1818 PAE support will be set to the previously hardcoded values, but the
1819 (previously hardcoded) path to the CDROM ISO image will be unset and
1820 if required, needs to be set manually with ``gnt-instance modify``
1822 - The address to which an instance's VNC console is bound is now
1823 selectable per-instance, rather than being cluster wide. Of course
1824 this only applies to instances controlled via VNC, so currently just
1825 applies to HVM clusters.
1831 *(Released Mon, 18 Feb 2008)*
1833 - more tweaks to the disk activation code (especially helpful for DRBD)
1834 - change the default ``gnt-instance list`` output format, now there is
1835 one combined status field (see the manpage for the exact values this
1837 - some more fixes for the mac export to hooks change
1838 - make Ganeti not break with DRBD 8.2.x (which changed the version
1839 format in ``/proc/drbd``) (issue 24)
1840 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1841 template, allowing migration from DRBD0.7+MD to DRBD8
1847 *(Released Wed, 30 Jan 2008)*
1849 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1851 - add command aliases infrastructure and a few aliases
1852 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1853 man pages and the ``--help`` option of ``gnt-node
1854 list``/``gnt-instance list``
1855 - fix ``gnt-backup list`` with down nodes (issue 21)
1856 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1857 - fix the dist archive and add a check for including svn/git files in
1859 - some developer-related changes: improve the burnin and the QA suite,
1860 add an upload script for testing during development
1866 *(Released Wed, 16 Jan 2008)*
1868 - experimental HVM support, read the install document, section
1869 "Initializing the cluster"
1870 - allow for the PVM hypervisor per-instance kernel and initrd paths
1871 - add a new command ``gnt-cluster verify-disks`` which uses a new
1872 algorithm to improve the reconnection of the DRBD pairs if the device
1873 on the secondary node has gone away
1874 - make logical volume code auto-activate LVs at disk activation time
1875 - slightly improve the speed of activating disks
1876 - allow specification of the MAC address at instance creation time, and
1877 changing it later via ``gnt-instance modify``
1878 - fix handling of external commands that generate lots of output on
1880 - update documentation with regard to minimum version of DRBD8 supported
1886 *(Released Tue, 4 Dec 2007)*
1888 - Log the ``xm create`` output to the node daemon log on failure (to
1889 help diagnosing the error)
1890 - In debug mode, log all external commands output if failed to the logs
1891 - Change parsing of lvm commands to ignore stderr
1897 *(Released Wed, 28 Nov 2007)*
1899 - Another round of updates to the DRBD 8 code to deal with more failures
1900 in the replace secondary node operation
1901 - Some more logging of failures in disk operations (lvm, drbd)
1902 - A few documentation updates
1909 *(Released Tue, 13 Nov 2007)*
1911 - Change configuration file format from Python's Pickle to JSON.
1912 Upgrading is possible using the cfgupgrade utility.
1913 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1914 faster replace disks and is more stable (DRBD 8 has many improvements
1915 compared to DRBD 0.7)
1916 - Added command line tags support (see man pages for ``gnt-instance``,
1917 ``gnt-node``, ``gnt-cluster``)
1918 - Added instance rename support
1919 - Added multi-instance startup/shutdown
1920 - Added cluster rename support
1921 - Added ``gnt-node evacuate`` to simplify some node operations
1922 - Added instance reboot operation that can speedup reboot as compared to
1924 - Soften the requirement that hostnames are in FQDN format
1925 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1927 - Removed dependency on debian's patched fping that uses the
1928 non-standard ``-S`` option
1929 - Now the OS definitions are searched for in multiple, configurable
1930 paths (easier for distros to package)
1931 - Some changes to the hooks infrastructure (especially the new
1932 post-configuration update hook)
1933 - Other small bugfixes
1935 .. vim: set textwidth=72 syntax=rst :