9 Incompatible/important changes and bugfixes
10 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
12 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
13 resource's ``ip_check`` parameter changed from ``True`` to ``False``
14 to match the underlying LUXI interface.
15 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
16 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
17 not maintain backwards-compatibility as the underlying operation
18 changed in an incompatible way. The RAPI client can talk to old
19 servers, but it needs to be told so as the return value changed.
20 - When creating file-based instances via RAPI, the ``file_driver``
21 parameter no longer defaults to ``loop`` and must be specified.
22 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
24 - Support for the undocumented and deprecated RAPI instance creation
25 request format version 0 has been dropped. Use version 1, supported
26 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
27 - Pyparsing 1.4.6 or above is required, see :doc:`installation
28 documentation <install>`.
29 - The "cluster-verify" hooks are now executed per group by the
30 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
31 if you just run ``gnt-cluster verify``, which generates one opcode per
33 - The environment as passed to the OS scripts is cleared, and thus no
34 environment variables defined in the node daemon's environment will be
35 inherited by the scripts.
36 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
38 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
39 support operations involving multiple node groups.
40 - Offline nodes are ignored when failing over an instance.
45 - The ganeti-htools project has been merged into the ganeti-core source
46 tree and will be built as part of Ganeti (see :doc:`install-quick`).
47 - Implemented support for :doc:`shared storage <design-shared-storage>`.
48 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
49 GPT-style partition tables (requires `parted
50 <http://www.gnu.org/s/parted/>`_).
51 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
52 - Allowed adding tags on instance creation.
53 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
54 - Allow instances to be started in a paused state, enabling the user to
55 see the complete console output on boot using the console.
56 - Added new hypervisor flag to control default reboot behaviour
57 (``reboot_behavior``).
58 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
59 - Improved out-of-band management support:
61 - Added ``gnt-node health`` command reporting the health status of
63 - Added ``gnt-node power`` command to manage power status of nodes.
64 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
66 - Instance migration can fall back to failover if instance is not
68 - Filters can be used when listing nodes, instances, groups and locks;
69 see *ganeti(7)* manpage.
70 - Added post-execution status as variables to :doc:`hooks <hooks>`
72 - Instance tags are exported/imported together with the instance.
73 - When given an explicit job ID, ``gnt-job info`` will work for archived
75 - Jobs can define dependencies on other jobs (not yet supported via
76 RAPI or command line, but used by internal commands and usable via
79 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
82 - Instance failover is now available as a RAPI resource
83 (``/2/instances/[instance_name]/failover``).
84 - ``gnt-instance info`` defaults to static information if primary node
86 - Opcodes have a new ``comment`` attribute.
88 Node group improvements
89 ~~~~~~~~~~~~~~~~~~~~~~~
91 - ``gnt-cluster verify`` has been modified to check groups separately,
92 thereby improving performance.
93 - Node group support has been added to ``gnt-cluster verify-disks``,
94 which now operates per node group.
95 - Watcher has been changed to work better with node groups.
97 - One process and state file per node group.
98 - Slow watcher in one group doesn't block other group's watcher.
100 - Added new command, ``gnt-group evacuate``, to move all instances in a
101 node group to other groups.
102 - Added ``gnt-instance change-group`` to move an instance to another
104 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
105 per-group operations.
106 - Node groups can be tagged.
107 - Some operations switch from an exclusive to a shared lock as soon as
109 - Instance's primary and secondary nodes' groups are now available as
110 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
111 and ``snodes.group.uuid``).
116 - Numerous updates to documentation and manpages.
118 - :doc:`RAPI <rapi>` documentation now has detailed parameter
121 - A lockset's internal lock is now also visible in lock monitor.
122 - Log messages from job queue workers now contain information about the
123 opcode they're processing.
124 - ``gnt-instance console`` no longer requires the instance lock.
125 - A short delay when waiting for job changes reduces the number of LUXI
126 requests significantly.
127 - DRBD metadata volumes are overwritten with zeros during disk creation.
133 *(Released Fri, 5 Aug 2011)*
135 Many bug-fixes and a few small features:
137 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
138 issues when you tried to add an instance with two MAC addresses in one
140 - KVM: fixed per-instance stored UID value
141 - KVM: configure bridged NICs at migration start
142 - KVM: Fix a bug where instance will not start with never KVM versions
144 - Added OS search path to ``gnt-cluster info``
145 - Fixed an issue with ``file_storage_dir`` where you were forced to
146 provide an absolute path, but the documentation states it is a
147 relative path, the documentation was right
148 - Added a new parameter to instance stop/start called ``--no-remember``
149 that will make the state change to not be remembered
150 - Implemented ``no_remember`` at RAPI level
151 - Improved the documentation
152 - Node evacuation: don't call IAllocator if node is already empty
153 - Fixed bug in DRBD8 replace disks on current nodes
154 - Fixed bug in recreate-disks for DRBD instances
155 - Moved assertion checking locks in ``gnt-instance replace-disks``
156 causing it to abort with not owning the right locks for some situation
157 - Job queue: Fixed potential race condition when cancelling queued jobs
158 - Fixed off-by-one bug in job serial generation
159 - ``gnt-node volumes``: Fix instance names
160 - Fixed aliases in bash completion
161 - Fixed a bug in reopening log files after beeing sent a SIGHUP
162 - Added a flag to burnin to allow specifying VCPU count
163 - Bugfixes to non-root Ganeti configuration
169 *(Released Thu, 12 May 2011)*
171 Many bug-fixes and a few new small features:
173 - Fixed a bug related to log opening failures
174 - Fixed a bug in instance listing with orphan instances
175 - Fixed a bug which prevented resetting the cluster-level node parameter
176 ``oob_program`` to the default
177 - Many fixes related to the ``cluster-merge`` tool
178 - Fixed a race condition in the lock monitor, which caused failures
179 during (at least) creation of many instances in parallel
180 - Improved output for gnt-job info
181 - Removed the quiet flag on some ssh calls which prevented debugging
183 - Improved the N+1 failure messages in cluster verify by actually
184 showing the memory values (needed and available)
185 - Increased lock attempt timeouts so that when executing long operations
186 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
187 too early and thus prevent the use of the 'fair' mechanism
188 - Changed instance query data (``gnt-instance info``) to not acquire
189 locks unless needed, thus allowing its use on locked instance if only
190 static information is asked for
191 - Improved behaviour with filesystems that do not support rename on an
193 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
194 kept locks on all nodes during the wipe, which is unneeded
195 - Fixed ``gnt-watcher`` handling of errors during hooks execution
196 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
197 10GiB) which caused the wipe to fail right at the end in some cases
198 - Fixed master IP activation when doing master failover with no-voting
199 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
200 the master node itself
201 - Fixed potential data-loss in under disk full conditions, where Ganeti
202 wouldn't check correctly the return code and would consider
203 partially-written files 'correct'
204 - Fixed bug related to multiple VGs and DRBD disk replacing
205 - Added new disk parameter ``metavg`` that allows placement of the meta
206 device for DRBD in a different volume group
207 - Fixed error handling in the node daemon when the system libc doesn't
208 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
209 - Fixed lock release during replace-disks, which kept cluster-wide locks
210 when doing disk replaces with an iallocator script
211 - Added check for missing bridges in cluster verify
212 - Handle EPIPE errors while writing to the terminal better, so that
213 piping the output to e.g. ``less`` doesn't cause a backtrace
214 - Fixed rare case where a ^C during Luxi calls could have been
215 interpreted as server errors, instead of simply terminating
216 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
218 - Added a few more parameters to the KVM hypervisor, allowing a second
219 CDROM, custom disk type for CDROMs and a floppy image
220 - Removed redundant message in instance rename when the name is given
222 - Added option to ``gnt-instance recreate-disks`` to allow creating the
223 disks on new nodes, allowing recreation when the original instance
224 nodes are completely gone
225 - Added option when converting disk templates to DRBD to skip waiting
226 for the resync, in order to make the instance available sooner
227 - Added two new variables to the OS scripts environment (containing the
229 - Made the root_path and optional parameter for the xen-pvm hypervisor,
230 to allow use of ``pvgrub`` as bootloader
231 - Changed the instance memory modifications to only check out-of-memory
232 conditions on memory increases, and turned the secondary node warnings
233 into errors (they can still be overridden via ``--force``)
234 - Fixed the handling of a corner case when the Python installation gets
235 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
236 to execute a command that doesn't exist
237 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
238 node of the instance returned failures during instance shutdown; this
239 adds the option ``--ignore-consistency`` to gnt-instance move
241 And as usual, various improvements to the error messages, documentation
247 *(Released Wed, 09 Mar 2011)*
249 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
250 the RAPI users file if run twice (even with ``--dry-run``).
252 The release fixes that bug (nothing else changed).
258 *(Released Mon, 07 Mar 2011)*
260 Final 2.4.0 release. Just a few small fixes:
262 - Fixed RAPI node evacuate
263 - Fixed the kvm-ifup script
264 - Fixed internal error handling for special job cases
265 - Updated man page to specify the escaping feature for options
271 *(Released Mon, 28 Feb 2011)*
273 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
274 that this feature wiped the disks of the wrong instance, leading to loss
279 - Fixed title of query field containing instance name
280 - Expanded the glossary in the documentation
281 - Fixed one unittest (internal issue)
287 *(Released Mon, 21 Feb 2011)*
289 A number of bug fixes plus just a couple functionality changes.
291 On the user-visible side, the ``gnt-* list`` command output has changed
292 with respect to "special" field states. The current rc1 style of display
293 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
294 the default output mode special fields states are displayed as follows:
296 - Offline resource: ``*``
297 - Unavailable/not applicable: ``-``
298 - Data missing (RPC failure): ``?``
299 - Unknown field: ``??``
301 Another user-visible change is the addition of ``--force-join`` to
306 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
307 - Fixed regression in RAPI/instance reinstall where all parameters were
308 required (instead of optional)
309 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
310 - Fixed iallocator usage (offline nodes were not considered offline)
311 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
312 - Fixed hypervisor and OS parameter validation with respect to
314 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
316 - Fixed ``tools/listrunner`` with respect to agent-based usage
322 *(Released Fri, 4 Feb 2011)*
324 Many changes and fixes since the beta1 release. While there were some
325 internal changes, the code has been mostly stabilised for the RC
328 Note: the dumb allocator was removed in this release, as it was not kept
329 up-to-date with the IAllocator protocol changes. It is recommended to
330 use the ``hail`` command from the ganeti-htools package.
332 Note: the 2.4 and up versions of Ganeti are not compatible with the
333 0.2.x branch of ganeti-htools. You need to upgrade to
334 ganeti-htools-0.3.0 (or later).
336 Regressions fixed from 2.3
337 ~~~~~~~~~~~~~~~~~~~~~~~~~~
339 - Fixed the ``gnt-cluster verify-disks`` command
340 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
342 - Fixed disk adoption breakage
343 - Fixed wrong headers in instance listing for field aliases
348 - Fixed corner case in KVM handling of NICs
349 - Fixed many cases of wrong handling of non-vm_capable nodes
350 - Fixed a bug where a missing instance symlink was not possible to
351 recreate with any ``gnt-*`` command (now ``gnt-instance
352 activate-disks`` does it)
353 - Fixed the volume group name as reported by ``gnt-cluster
355 - Increased timeouts for the import-export code, hopefully leading to
356 fewer aborts due network or instance timeouts
357 - Fixed bug in ``gnt-node list-storage``
358 - Fixed bug where not all daemons were started on cluster
359 initialisation, but only at the first watcher run
360 - Fixed many bugs in the OOB implementation
361 - Fixed watcher behaviour in presence of instances with offline
363 - Fixed instance list output for instances running on the wrong node
364 - a few fixes to the cluster-merge tool, but it still cannot merge
365 multi-node groups (currently it is not recommended to use this tool)
371 - Improved network configuration for the KVM hypervisor
372 - Added e1000 as a supported NIC for Xen-HVM
373 - Improved the lvmstrap tool to also be able to use partitions, as
374 opposed to full disks
375 - Improved speed of disk wiping (the cluster parameter
376 ``prealloc_wipe_disks``, so that it has a low impact on the total time
377 of instance creations
378 - Added documentation for the OS parameters
379 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
380 hypervisor is not responding
381 - Added display of blacklisted and hidden OS information in
383 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
384 NIC and node parameters, which might create problems with currently
385 invalid (but undetected) configuration files, but prevents validation
386 failures when unrelated parameters are modified
387 - Changed cluster initialisation to wait for the master daemon to become
389 - Expanded the RAPI interface:
391 - Added config redistribution resource
392 - Added activation/deactivation of instance disks
393 - Added export of console information
395 - Implemented log file reopening on SIGHUP, which allows using
396 logrotate(8) for the Ganeti log files
397 - Added a basic OOB helper script as an example
403 *(Released Fri, 14 Jan 2011)*
408 - Fixed timezone issues when formatting timestamps
409 - Added support for node groups, available via ``gnt-group`` and other
411 - Added out-of-band framework and management, see :doc:`design
412 document <design-oob>`
413 - Removed support for roman numbers from ``gnt-node list`` and
414 ``gnt-instance list``.
415 - Allowed modification of master network interface via ``gnt-cluster
416 modify --master-netdev``
417 - Accept offline secondaries while shutting down instance disks
418 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
419 - Added support for multiple LVM volume groups
420 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
422 - Added commands to list available fields:
424 - ``gnt-node list-fields``
425 - ``gnt-group list-fields``
426 - ``gnt-instance list-fields``
428 - Updated documentation and man pages
433 - Moved ``rapi_users`` file into separate directory, now named
434 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
436 - Added new tool for running commands on many machines,
437 ``tools/ganeti-listrunner``
438 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
439 improving the ``gnt-instance console`` output
440 - Allowed customisation of disk index separator at ``configure`` time
441 - Export node group allocation policy to :doc:`iallocator <iallocator>`
442 - Added support for non-partitioned md disks in ``lvmstrap``
443 - Added script to gracefully power off KVM instances
444 - Split ``utils`` module into smaller parts
445 - Changed query operations to return more detailed information, e.g.
446 whether an information is unavailable due to an offline node. To use
447 this new functionality, the LUXI call ``Query`` must be used. Field
448 information is now stored by the master daemon and can be retrieved
449 using ``QueryFields``. Instances, nodes and groups can also be queried
450 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
451 exposed via RAPI). The following commands make use of this
452 infrastructure change:
455 - ``gnt-group list-fields``
457 - ``gnt-node list-fields``
458 - ``gnt-instance list``
459 - ``gnt-instance list-fields``
460 - ``gnt-debug locks``
465 - New RAPI resources (see :doc:`rapi`):
469 - ``/2/groups/[group_name]``
470 - ``/2/groups/[group_name]/assign-nodes``
471 - ``/2/groups/[group_name]/modify``
472 - ``/2/groups/[group_name]/rename``
473 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
477 - Implemented ``no_install`` for instance creation
478 - Implemented OS parameters for instance reinstallation, allowing
479 use of special settings on reinstallation (e.g. for preserving data)
484 - Added IPv6 support in import/export
485 - Pause DRBD synchronization while wiping disks on instance creation
486 - Updated unittests and QA scripts
487 - Improved network parameters passed to KVM
488 - Converted man pages from docbook to reStructuredText
494 *(Released Mon, 20 Dec 2010)*
496 Released version 2.3.1~rc1 without any changes.
502 *(Released Wed, 1 Dec 2010)*
504 - impexpd: Disable OpenSSL compression in socat if possible (backport
505 from master, commit e90739d625b, see :doc:`installation guide
506 <install-quick>` for details)
507 - Changed unittest coverage report to exclude test scripts
508 - Added script to check version format
514 *(Released Wed, 1 Dec 2010)*
516 Released version 2.3.0~rc1 without any changes.
522 *(Released Fri, 19 Nov 2010)*
524 A number of bugfixes and documentation updates:
526 - Update ganeti-os-interface documentation
527 - Fixed a bug related to duplicate MACs or similar items which should be
529 - Fix breakage in OS state modify
530 - Reinstall instance: disallow offline secondaries (fixes bug related to
531 OS changing but reinstall failing)
532 - plus all the other fixes between 2.2.1 and 2.2.2
538 *(Released Tue, 2 Nov 2010)*
540 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
541 - Fixed master failover race with watcher
542 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
544 - Accept previously stopped instance for export with instance removal
545 - Simplify and extend the environment variables for instance OS scripts
546 - Added new node flags, ``master_capable`` and ``vm_capable``
547 - Added optional instance disk wiping prior during allocation. This is a
548 cluster-wide option and can be set/modified using
549 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
550 - Added IPv6 support, see :doc:`design document <design-2.3>` and
552 - Added a new watcher option (``--ignore-pause``)
553 - Added option to ignore offline node on instance start/stop
554 (``--ignore-offline``)
555 - Allow overriding OS parameters with ``gnt-instance reinstall``
556 - Added ability to change node's secondary IP address using ``gnt-node
558 - Implemented privilege separation for all daemons except
559 ``ganeti-noded``, see ``configure`` options
560 - Complain if an instance's disk is marked faulty in ``gnt-cluster
562 - Implemented job priorities (see ``ganeti(7)`` manpage)
563 - Ignore failures while shutting down instances during failover from
565 - Exit daemon's bootstrap process only once daemon is ready
566 - Export more information via ``LUInstanceQuery``/remote API
567 - Improved documentation, QA and unittests
568 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
569 restart if the file was created or changed
570 - Added LUXI protocol version sent with each request and response,
571 allowing detection of server/client mismatches
572 - Moved the Python scripts among gnt-* and ganeti-* into modules
573 - Moved all code related to setting up SSH to an external script,
575 - Infrastructure changes for node group support in future versions
581 *(Released Fri, 19 Nov 2010)*
583 A few small bugs fixed, and some improvements to the build system:
585 - Fix documentation regarding conversion to drbd
586 - Fix validation of parameters in cluster modify (``gnt-cluster modify
588 - Fix error handling in node modify with multiple changes
589 - Allow remote imports without checked names
595 *(Released Tue, 19 Oct 2010)*
597 - Disable SSL session ID cache in RPC client
603 *(Released Thu, 14 Oct 2010)*
605 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
606 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
608 - Fix problems with interaction between Curl and Python's HTTP server,
609 resulting in increased speed in many RPC calls
610 - Improve our release script to prevent breakage with older aclocal and
617 *(Released Thu, 7 Oct 2010)*
619 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
620 value retrieved from master
621 - Added support for blacklisted or hidden OS definitions
622 - Added simple lock monitor (accessible via (``gnt-debug locks``)
623 - Added support for -mem-path in KVM hypervisor abstraction layer
624 - Allow overriding instance parameters in tool for inter-cluster
625 instance moves (``tools/move-instance``)
626 - Improved opcode summaries (e.g. in ``gnt-job list``)
627 - Improve consistency of OS listing by sorting it
628 - Documentation updates
634 *(Released Fri, 8 Oct 2010)*
636 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
642 *(Released Mon, 4 Oct 2010)*
644 - Fixed regression in ``gnt-instance rename``
650 *(Released Wed, 22 Sep 2010)*
652 - Fixed OS_VARIANT variable for OS scripts
653 - Fixed cluster tag operations via RAPI
654 - Made ``setup-ssh`` exit with non-zero code if an error occurred
655 - Disabled RAPI CA checks in watcher
661 *(Released Mon, 23 Aug 2010)*
663 - Support DRBD versions of the format "a.b.c.d"
665 - Re-introduce support for usage from multiple threads in RAPI client
666 - Instance renames and modify via RAPI
667 - Work around race condition between processing and archival in job
669 - Mark opcodes following failed one as failed, too
670 - Job field ``lock_status`` was removed due to difficulties making it
671 work with the changed job queue in Ganeti 2.2; a better way to monitor
672 locks is expected for a later 2.2.x release
673 - Fixed dry-run behaviour with many commands
674 - Support ``ssh-agent`` again when adding nodes
675 - Many additional bugfixes
681 *(Released Fri, 30 Jul 2010)*
683 Important change: the internal RPC mechanism between Ganeti nodes has
684 changed from using a home-grown http library (based on the Python base
685 libraries) to use the PycURL library. This requires that PycURL is
686 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
687 against GnuTLS by default. cURL's support for GnuTLS had known issues
688 before cURL 7.21.0 and we recommend using the latest cURL release or
689 linking against OpenSSL. Most other distributions already link PycURL
690 and cURL against OpenSSL. The command::
692 python -c 'import pycurl; print pycurl.version'
694 can be used to determine the libraries PycURL and cURL are linked
697 Other significant changes:
699 - Rewrote much of the internals of the job queue, in order to achieve
700 better parallelism; this decouples job query operations from the job
701 processing, and it should allow much nicer behaviour of the master
702 daemon under load, and it also has uncovered some long-standing bugs
703 related to the job serialisation (now fixed)
704 - Added a default iallocator setting to the cluster parameters,
705 eliminating the need to always pass nodes or an iallocator for
706 operations that require selection of new node(s)
707 - Added experimental support for the LXC virtualization method
708 - Added support for OS parameters, which allows the installation of
709 instances to pass parameter to OS scripts in order to customise the
711 - Added a hypervisor parameter controlling the migration type (live or
712 non-live), since hypervisors have various levels of reliability; this
713 has renamed the 'live' parameter to 'mode'
714 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
715 logical volumes, meaning that cluster verify will ignore them and not
716 flag their presence as errors
717 - The watcher will now reset the error count for failed instances after
718 8 hours, thus allowing self-healing if the problem that caused the
719 instances to be down/fail to start has cleared in the meantime
720 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
721 check for, and warn, if the drbd module parameter ``usermode_helper``
722 is not consistent with the cluster-wide setting; this is needed to
723 make diagnose easier of failed drbd creations
724 - Started adding base IPv6 support, but this is not yet
725 enabled/available for use
726 - Rename operations (cluster, instance) will now return the new name,
727 which is especially useful if a short name was passed in
728 - Added support for instance migration in RAPI
729 - Added a tool to pre-configure nodes for the SSH setup, before joining
730 them to the cluster; this will allow in the future a simplified model
731 for node joining (but not yet fully enabled in 2.2); this needs the
732 paramiko python library
733 - Fixed handling of name-resolving errors
734 - Fixed consistency of job results on the error path
735 - Fixed master-failover race condition when executed multiple times in
737 - Fixed many bugs related to the job queue (mostly introduced during the
738 2.2 development cycle, so not all are impacting 2.1)
739 - Fixed instance migration with missing disk symlinks
740 - Fixed handling of unknown jobs in ``gnt-job archive``
741 - And many other small fixes/improvements
745 - Enhanced both the unittest and the QA coverage
746 - Switched the opcode validation to a generic model, and extended the
747 validation to all opcode parameters
748 - Changed more parts of the code that write shell scripts to use the
750 - Switched the master daemon to use the asyncore library for the Luxi
757 *(Released Thu, 17 Jun 2010)*
759 - Added tool (``move-instance``) and infrastructure to move instances
760 between separate clusters (see :doc:`separate documentation
761 <move-instance>` and :doc:`design document <design-2.2>`)
762 - Added per-request RPC timeout
763 - RAPI now requires a Content-Type header for requests with a body (e.g.
764 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
765 :rfc:`2616` (HTTP/1.1), section 7.2.1)
766 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
768 - Implemented initial support for running Ganeti daemons as separate
769 users, see configure-time flags ``--with-user-prefix`` and
770 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
772 - Instances can be removed after export (``gnt-backup export
774 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
775 key (instead of 1024 bit)
776 - Added new cluster configuration file for cluster domain secret
777 - Import/export now use SSL instead of SSH
778 - Added support for showing estimated time when exporting an instance,
779 see the ``ganeti-os-interface(7)`` manpage and look for
786 *(Released Tue, 16 Nov 2010)*
788 Some more bugfixes. Unless critical bugs occur, this will be the last
791 - Fix case of MAC special-values
792 - Fix mac checker regex
793 - backend: Fix typo causing "out of range" error
794 - Add missing --units in gnt-instance list man page
800 *(Released Tue, 24 Aug 2010)*
803 - Don't ignore secondary node silently on non-mirrored disk templates
805 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
806 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
807 - Properly document the usb_mouse parameter
808 - Fix path in ganeti-rapi(8) (issue 116)
809 - Adjust error message when the ganeti user's .ssh directory is
811 - Add same-node-check when changing the disk template to drbd
817 *(Released Fri, 16 Jul 2010)*
820 - Add an option to only select some reboot types during qa/burnin.
821 (on some hypervisors consequent reboots are not supported)
822 - Fix infrequent race condition in master failover. Sometimes the old
823 master ip address would be still detected as up for a short time
824 after it was removed, causing failover to fail.
825 - Decrease mlockall warnings when the ctypes module is missing. On
826 Python 2.4 we support running even if no ctypes module is installed,
827 but we were too verbose about this issue.
828 - Fix building on old distributions, on which man doesn't have a
830 - Fix RAPI not to ignore the MAC address on instance creation
831 - Implement the old instance creation format in the RAPI client.
837 *(Released Thu, 01 Jul 2010)*
839 A small bugfix release:
840 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
841 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
842 a non-existing option
843 - Fix instance create: the --force-variant option was ignored
844 - Improve pylint 0.21 compatibility and warnings with Python 2.6
845 - Fix modify node storage with non-FQDN arguments
846 - Fix RAPI client to authenticate under Python 2.6 when used
847 for more than 5 requests needing authentication
848 - Fix gnt-instance modify -t (storage) giving a wrong error message
849 when converting a non-shutdown drbd instance to plain
855 *(Released Fri, 18 Jun 2010)*
857 A small bugfix release:
859 - Fix live migration of KVM instances started with older Ganeti
860 versions which had fewer hypervisor parameters
861 - Fix gnt-instance grow-disk on down instances
862 - Fix an error-reporting bug during instance migration
863 - Better checking of the ``--net`` and ``--disk`` values, to avoid
864 silently ignoring broken ones
865 - Fix an RPC error reporting bug affecting, for example, RAPI client
867 - Fix bug triggered by different API version os-es on different nodes
868 - Fix a bug in instance startup with custom hvparams: OS level
869 parameters would fail to be applied.
870 - Fix the RAPI client under Python 2.6 (but more work is needed to
871 make it work completely well with OpenSSL)
872 - Fix handling of errors when resolving names from DNS
878 *(Released Thu, 3 Jun 2010)*
880 A medium sized development cycle. Some new features, and some
881 fixes/small improvements/cleanups.
886 The node deamon now tries to mlock itself into memory, unless the
887 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
888 its logs, and falls back to console logging. This allows emergency
889 features such as ``gnt-node powercycle`` to work even in the event of a
890 broken node disk (tested offlining the disk hosting the node's
891 filesystem and dropping its memory caches; don't try this at home)
893 KVM: add vhost-net acceleration support. It can be tested with a new
894 enough version of the kernel and of qemu-kvm.
896 KVM: Add instance chrooting feature. If you use privilege dropping for
897 your VMs you can also now force them to chroot to an empty directory,
898 before starting the emulated guest.
900 KVM: Add maximum migration bandwith and maximum downtime tweaking
901 support (requires a new-enough version of qemu-kvm).
903 Cluster verify will now warn if the master node doesn't have the master
906 Add a new (incompatible) instance creation request format to RAPI which
907 supports all parameters (previously only a subset was supported, and it
908 wasn't possible to extend the old format to accomodate all the new
909 features. The old format is still supported, and a client can check for
910 this feature, before using it, by checking for its presence in the
911 ``features`` RAPI resource.
913 Now with ancient latin support. Try it passing the ``--roman`` option to
914 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
915 (requires the python-roman module to be installed, in order to work).
920 As usual many internal code refactorings, documentation updates, and
923 - Lots of improvements and cleanups to the experimental Remote API
924 (RAPI) client library.
925 - A new unit test suite for the core daemon libraries.
926 - A fix to creating missing directories makes sure the umask is not
927 applied anymore. This enforces the same directory permissions
929 - Better handling terminating daemons with ctrl+c (used when running
930 them in debugging mode).
931 - Fix a race condition in live migrating a KVM instance, when stat()
932 on the old proc status file returned EINVAL, which is an unexpected
934 - Fixed manpage checking with newer man and utf-8 charachters. But now
935 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
941 *(Released Fri, 7 May 2010)*
943 Fix a bug which prevented untagged KVM instances from starting.
949 *(Released Fri, 7 May 2010)*
951 Another release with a long development cycle, during which many
952 different features were added.
957 The KVM hypervisor now can run the individual instances as non-root, to
958 reduce the impact of a VM being hijacked due to bugs in the
959 hypervisor. It is possible to run all instances as a single (non-root)
960 user, to manually specify a user for each instance, or to dynamically
961 allocate a user out of a cluster-wide pool to each instance, with the
962 guarantee that no two instances will run under the same user ID on any
965 An experimental RAPI client library, that can be used standalone
966 (without the other Ganeti libraries), is provided in the source tree as
967 ``lib/rapi/client.py``. Note this client might change its interface in
968 the future, as we iterate on its capabilities.
970 A new command, ``gnt-cluster renew-crypto`` has been added to easily
971 replace the cluster's certificates and crypto keys. This might help in
972 case they have been compromised, or have simply expired.
974 A new disk option for instance creation has been added that allows one
975 to "adopt" currently existing logical volumes, with data
976 preservation. This should allow easier migration to Ganeti from
977 unmanaged (or managed via other software) instances.
979 Another disk improvement is the possibility to convert between redundant
980 (DRBD) and plain (LVM) disk configuration for an instance. This should
981 allow better scalability (starting with one node and growing the
982 cluster, or shrinking a two-node cluster to one node).
984 A new feature that could help with automated node failovers has been
985 implemented: if a node sees itself as offline (by querying the master
986 candidates), it will try to shutdown (hard) all instances and any active
987 DRBD devices. This reduces the risk of duplicate instances if an
988 external script automatically failovers the instances on such nodes. To
989 enable this, the cluster parameter ``maintain_node_health`` should be
990 enabled; in the future this option (per the name) will enable other
991 automatic maintenance features.
993 Instance export/import now will reuse the original instance
994 specifications for all parameters; that means exporting an instance,
995 deleting it and the importing it back should give an almost identical
996 instance. Note that the default import behaviour has changed from
997 before, where it created only one NIC; now it recreates the original
1000 Cluster verify has added a few new checks: SSL certificates validity,
1001 /etc/hosts consistency across the cluster, etc.
1006 As usual, many internal changes were done, documentation fixes,
1009 - Fixed cluster initialization with disabled cluster storage (regression
1010 introduced in 2.1.1)
1011 - File-based storage supports growing the disks
1012 - Fixed behaviour of node role changes
1013 - Fixed cluster verify for some corner cases, plus a general rewrite of
1014 cluster verify to allow future extension with more checks
1015 - Fixed log spamming by watcher and node daemon (regression introduced
1017 - Fixed possible validation issues when changing the list of enabled
1019 - Fixed cleanup of /etc/hosts during node removal
1020 - Fixed RAPI response for invalid methods
1021 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1022 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1023 from ide disks, etc.)
1024 - Allow OS changes without re-installation (to record a changed OS
1025 outside of Ganeti, or to allow OS renames)
1026 - Allow instance creation without OS installation (useful for example if
1027 the OS will be installed manually, or restored from a backup not in
1029 - Implemented option to make cluster ``copyfile`` use the replication
1031 - Added list of enabled hypervisors to ssconf (possibly useful for
1033 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1035 - A partial form of node re-IP is possible via node readd, which now
1036 allows changed node primary IP
1037 - Command line utilities now show an informational message if the job is
1039 - The logs of the master daemon now show the PID/UID/GID of the
1046 *(Released Fri, 12 Mar 2010)*
1048 During the 2.1.0 long release candidate cycle, a lot of improvements and
1049 changes have accumulated with were released later as 2.1.1.
1054 The node evacuate command (``gnt-node evacuate``) was significantly
1055 rewritten, and as such the IAllocator protocol was changed - a new
1056 request type has been added. This unfortunate change during a stable
1057 series is designed to improve performance of node evacuations; on
1058 clusters with more than about five nodes and which are well-balanced,
1059 evacuation should proceed in parallel for all instances of the node
1060 being evacuated. As such, any existing IAllocator scripts need to be
1061 updated, otherwise the above command will fail due to the unknown
1062 request. The provided "dumb" allocator has not been updated; but the
1063 ganeti-htools package supports the new protocol since version 0.2.4.
1065 Another important change is increased validation of node and instance
1066 names. This might create problems in special cases, if invalid host
1067 names are being used.
1069 Also, a new layer of hypervisor parameters has been added, that sits at
1070 OS level between the cluster defaults and the instance ones. This allows
1071 customisation of virtualization parameters depending on the installed
1072 OS. For example instances with OS 'X' may have a different KVM kernel
1073 (or any other parameter) than the cluster defaults. This is intended to
1074 help managing a multiple OSes on the same cluster, without manual
1075 modification of each instance's parameters.
1077 A tool for merging clusters, ``cluster-merge``, has been added in the
1078 tools sub-directory.
1083 - Improved the int/float conversions that should make the code more
1084 robust in face of errors from the node daemons
1085 - Fixed the remove node code in case of internal configuration errors
1086 - Fixed the node daemon behaviour in face of inconsistent queue
1087 directory (e.g. read-only file-system where we can't open the files
1089 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1090 now it either aborts cleanly or, if given the new "auto_promote"
1091 parameter, will automatically promote other nodes as needed
1092 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1093 completely prevent Ganeti from working
1094 - Fixed bug for instance export when not all disks were successfully
1096 - Fixed behaviour of node add when the new node is slow in starting up
1098 - Fixed handling of signals in the LUXI client, which should improve
1099 behaviour of command-line scripts
1100 - Added checks for invalid node/instance names in the configuration (now
1101 flagged during cluster verify)
1102 - Fixed watcher behaviour for disk activation errors
1103 - Fixed two potentially endless loops in http library, which led to the
1104 RAPI daemon hanging and consuming 100% CPU in some cases
1105 - Fixed bug in RAPI daemon related to hashed passwords
1106 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1108 - Enhanced compatibility with non-Debian OSes, but not using absolute
1109 path in some commands and allowing customisation of the ssh
1110 configuration directory
1111 - Fixed possible future issue with new Python versions by abiding to the
1112 proper use of ``__slots__`` attribute on classes
1113 - Added checks that should prevent directory traversal attacks
1114 - Many documentation fixes based on feedback from users
1119 - Added an "early_release" more for instance replace disks and node
1120 evacuate, where we release locks earlier and thus allow higher
1121 parallelism within the cluster
1122 - Added watcher hooks, intended to allow the watcher to restart other
1123 daemons (e.g. from the ganeti-nbma project), but they can be used of
1124 course for any other purpose
1125 - Added a compile-time disable for DRBD barriers, to increase
1126 performance if the administrator trusts the power supply or the
1127 storage system to not lose writes
1128 - Added the option of using syslog for logging instead of, or in
1129 addition to, Ganeti's own log files
1130 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1131 can indeed boot from a paravirtual NIC
1132 - Added a generic debug level for many operations; while this is not
1133 used widely yet, it allows one to pass the debug value all the way to
1135 - Enhanced the hooks environment for instance moves (failovers,
1136 migrations) where the primary/secondary nodes changed during the
1137 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1138 - Enhanced data validations for many user-supplied values; one important
1139 item is the restrictions imposed on instance and node names, which
1140 might reject some (invalid) host names
1141 - Add a configure-time option to disable file-based storage, if it's not
1142 needed; this allows greater security separation between the master
1143 node and the other nodes from the point of view of the inter-node RPC
1145 - Added user notification in interactive tools if job is waiting in the
1146 job queue or trying to acquire locks
1147 - Added log messages when a job is waiting for locks
1148 - Added filtering by node tags in instance operations which admit
1149 multiple instances (start, stop, reboot, reinstall)
1150 - Added a new tool for cluster mergers, ``cluster-merge``
1151 - Parameters from command line which are of the form ``a=b,c=d`` can now
1152 use backslash escapes to pass in values which contain commas,
1153 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1155 - For KVM, the instance name is the first parameter passed to KVM, so
1156 that it's more visible in the process list
1162 *(Released Tue, 2 Mar 2010)*
1164 Ganeti 2.1 brings many improvements with it. Major changes:
1166 - Added infrastructure to ease automated disk repairs
1167 - Added new daemon to export configuration data in a cheaper way than
1168 using the remote API
1169 - Instance NICs can now be routed instead of being associated with a
1171 - Improved job locking logic to reduce impact of jobs acquiring multiple
1172 locks waiting for other long-running jobs
1174 In-depth implementation details can be found in the Ganeti 2.1 design
1180 - Added chroot hypervisor
1181 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1183 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1184 ``bootloader_path`` and ``bootloader_args``)
1185 - Added the ``use_localtime`` option for the xen-hvm and kvm
1186 hypervisors, and the default value for this has changed to false (in
1187 2.0 xen-hvm always enabled it)
1188 - Added luxi call to submit multiple jobs in one go
1189 - Added cluster initialization option to not modify ``/etc/hosts``
1191 - Added network interface parameters
1192 - Added dry run mode to some LUs
1193 - Added RAPI resources:
1195 - ``/2/instances/[instance_name]/info``
1196 - ``/2/instances/[instance_name]/replace-disks``
1197 - ``/2/nodes/[node_name]/evacuate``
1198 - ``/2/nodes/[node_name]/migrate``
1199 - ``/2/nodes/[node_name]/role``
1200 - ``/2/nodes/[node_name]/storage``
1201 - ``/2/nodes/[node_name]/storage/modify``
1202 - ``/2/nodes/[node_name]/storage/repair``
1204 - Added OpCodes to evacuate or migrate all instances on a node
1205 - Added new command to list storage elements on nodes (``gnt-node
1206 list-storage``) and modify them (``gnt-node modify-storage``)
1207 - Added new ssconf files with master candidate IP address
1208 (``ssconf_master_candidates_ips``), node primary IP address
1209 (``ssconf_node_primary_ips``) and node secondary IP address
1210 (``ssconf_node_secondary_ips``)
1211 - Added ``ganeti-confd`` and a client library to query the Ganeti
1212 configuration via UDP
1213 - Added ability to run hooks after cluster initialization and before
1215 - Added automatic mode for disk replace (``gnt-instance replace-disks
1217 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1218 after catastrophic data-loss
1219 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1221 - Added ``gnt-instance move`` command to move instances
1222 - Added ``gnt-cluster watcher`` command to control watcher
1223 - Added ``gnt-node powercycle`` command to powercycle nodes
1224 - Added new job status field ``lock_status``
1225 - Added parseable error codes to cluster verification (``gnt-cluster
1226 verify --error-codes``) and made output less verbose (use
1227 ``--verbose`` to restore previous behaviour)
1228 - Added UUIDs to the main config entities (cluster, nodes, instances)
1229 - Added support for OS variants
1230 - Added support for hashed passwords in the Ganeti remote API users file
1232 - Added option to specify maximum timeout on instance shutdown
1233 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1234 - Added new helper script to start and stop Ganeti daemons
1235 (``daemon-util``), with the intent to reduce the work necessary to
1236 adjust Ganeti for non-Debian distributions and to start/stop daemons
1238 - Added more unittests
1239 - Fixed critical bug in ganeti-masterd startup
1240 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1241 now customisable at the cluster level for both the KVM and Xen
1242 hypervisors using the new ``migration_port`` parameter
1243 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1244 reinstalling an instance
1245 - Allowed ``@`` in tag names
1246 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1247 - Many documentation updates
1248 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1249 - ``gnt-instance reinstall`` can now reinstall multiple instances
1250 - Updated many command line parameters
1251 - Introduced new OS API version 15
1252 - No longer support a default hypervisor
1253 - Treat virtual LVs as inexistent
1254 - Improved job locking logic to reduce lock contention
1255 - Match instance and node names case insensitively
1256 - Reimplemented bash completion script to be more complete
1263 *(Released Thu, 4 Feb 2010)*
1265 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1266 - Fix a string formatting bug
1267 - Improve safety of the code in some error paths
1268 - Improve data validation in the master of values returned from nodes
1274 *(Released Thu, 17 Dec 2009)*
1276 - Fix security issue due to missing validation of iallocator names; this
1277 allows local and remote execution of arbitrary executables
1278 - Fix failure of gnt-node list during instance removal
1279 - Ship the RAPI documentation in the archive
1285 *(Released Wed, 30 Sep 2009)*
1287 - Fixed many wrong messages
1288 - Fixed a few bugs related to the locking library
1289 - Fixed MAC checking at instance creation time
1290 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1291 - Fixed a few issues related to signal handling in both daemons and
1293 - Fixed the example startup script provided
1294 - Fixed insserv dependencies in the example startup script (patch from
1296 - Fixed handling of drained nodes in the iallocator framework
1297 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1299 - Fixed error related to invalid job IDs in job polling
1300 - Fixed job/opcode persistence on unclean master shutdown
1301 - Fixed handling of partial job processing after unclean master
1303 - Fixed error reporting from LUs, previously all errors were converted
1304 into execution errors
1305 - Fixed error reporting from burnin
1306 - Decreased significantly the memory usage of the job queue
1307 - Optimised slightly multi-job submission
1308 - Optimised slightly opcode loading
1309 - Backported the multi-job submit framework from the development
1310 branch; multi-instance start and stop should be faster
1311 - Added script to clean archived jobs after 21 days; this will reduce
1312 the size of the queue directory
1313 - Added some extra checks in disk size tracking
1314 - Added an example ethers hook script
1315 - Added a cluster parameter that prevents Ganeti from modifying of
1317 - Added more node information to RAPI responses
1318 - Added a ``gnt-job watch`` command that allows following the ouput of a
1320 - Added a bind-address option to ganeti-rapi
1321 - Added more checks to the configuration verify
1322 - Enhanced the burnin script such that some operations can be retried
1324 - Converted instance reinstall to multi-instance model
1330 *(Released Fri, 7 Aug 2009)*
1332 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1333 to allow using the pre-2.0.2 behaviour in activation, if any existing
1334 instances have mismatched disk sizes in the configuration
1335 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1336 any configuration mismatches for disk sizes
1337 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1338 failover to work on two-node clusters
1339 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1341 - Fixed detection of OS script errors in ``gnt-backup export``
1342 - Fixed exit code of ``gnt-backup export``
1348 *(Released Fri, 17 Jul 2009)*
1350 - Added experimental support for stripped logical volumes; this should
1351 enhance performance but comes with a higher complexity in the block
1352 device handling; stripping is only enabled when passing
1353 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1354 affected even in the non-stripped mode
1355 - Improved resiliency against transient failures at the end of DRBD
1356 resyncs, and in general of DRBD resync checks
1357 - Fixed a couple of issues with exports and snapshot errors
1358 - Fixed a couple of issues in instance listing
1359 - Added display of the disk size in ``gnt-instance info``
1360 - Fixed checking for valid OSes in instance creation
1361 - Fixed handling of the "vcpus" parameter in instance listing and in
1362 general of invalid parameters
1363 - Fixed http server library, and thus RAPI, to handle invalid
1364 username/password combinations correctly; this means that now they
1365 report unauthorized for queries too, not only for modifications,
1366 allowing earlier detect of configuration problems
1367 - Added a new "role" node list field, equivalent to the master/master
1368 candidate/drained/offline flags combinations
1369 - Fixed cluster modify and changes of candidate pool size
1370 - Fixed cluster verify error messages for wrong files on regular nodes
1371 - Fixed a couple of issues with node demotion from master candidate role
1372 - Fixed node readd issues
1373 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1374 - Added a new ``--no-voting`` option for masterfailover to fix failover
1375 on two-nodes clusters when the former master node is unreachable
1376 - Added instance reinstall over RAPI
1382 *(Released Tue, 16 Jun 2009)*
1384 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1385 allow re-adding the start in single-user option (regression from 1.2)
1386 - the watcher writes the instance status to a file, to allow monitoring
1387 to report the instance status (from the master) based on cached
1388 results of the watcher's queries; while this can get stale if the
1389 watcher is being locked due to other work on the cluster, this is
1390 still an improvement
1391 - the watcher now also restarts the node daemon and the rapi daemon if
1393 - fixed the watcher to handle full and drained queue cases
1394 - hooks export more instance data in the environment, which helps if
1395 hook scripts need to take action based on the instance's properties
1396 (no longer need to query back into ganeti)
1397 - instance failovers when the instance is stopped do not check for free
1398 RAM, so that failing over a stopped instance is possible in low memory
1400 - rapi uses queries for tags instead of jobs (for less job traffic), and
1401 for cluster tags it won't talk to masterd at all but read them from
1403 - a couple of error handling fixes in RAPI
1404 - drbd handling: improved the error handling of inconsistent disks after
1405 resync to reduce the frequency of "there are some degraded disks for
1406 this instance" messages
1407 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1408 error handling path called a wrong function name)
1414 *(Released Wed, 27 May 2009)*
1416 - no changes from rc5
1419 Version 2.0 release candidate 5
1420 -------------------------------
1422 *(Released Wed, 20 May 2009)*
1424 - fix a couple of bugs (validation, argument checks)
1425 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1426 - some small improvements to RAPI and IAllocator
1427 - make watcher automatically start the master daemon if down
1430 Version 2.0 release candidate 4
1431 -------------------------------
1433 *(Released Mon, 27 Apr 2009)*
1435 - change the OS list to not require locks; this helps with big clusters
1436 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1437 volume group is broken
1438 - ``gnt-instance info``, without any arguments, doesn't run for all
1439 instances anymore; either pass ``--all`` or pass the desired
1440 instances; this helps against mistakes on big clusters where listing
1441 the information for all instances takes a long time
1442 - miscellaneous doc and man pages fixes
1445 Version 2.0 release candidate 3
1446 -------------------------------
1448 *(Released Wed, 8 Apr 2009)*
1450 - Change the internal locking model of some ``gnt-node`` commands, in
1451 order to reduce contention (and blocking of master daemon) when
1452 batching many creation/reinstall jobs
1453 - Fixes to Xen soft reboot
1454 - No longer build documentation at build time, instead distribute it in
1455 the archive, in order to reduce the need for the whole docbook/rst
1459 Version 2.0 release candidate 2
1460 -------------------------------
1462 *(Released Fri, 27 Mar 2009)*
1464 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1465 - Fix watcher startup sequence, improves the behaviour of busy clusters
1466 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1467 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1468 bind address and other places
1469 - Some documentation fixes and updates
1472 Version 2.0 release candidate 1
1473 -------------------------------
1475 *(Released Mon, 2 Mar 2009)*
1477 - More documentation updates, now all docs should be more-or-less
1479 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1481 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1483 - Fix an issue related to $libdir/run/ganeti and cluster creation
1489 *(Released Thu, 19 Feb 2009)*
1491 - Xen PVM and KVM have switched the default value for the instance root
1492 disk to the first partition on the first drive, instead of the whole
1493 drive; this means that the OS installation scripts must be changed
1495 - Man pages have been updated
1496 - RAPI has been switched by default to HTTPS, and the exported functions
1497 should all work correctly
1498 - RAPI v1 has been removed
1499 - Many improvements to the KVM hypervisor
1500 - Block device errors are now better reported
1501 - Many other bugfixes and small improvements
1507 *(Released Mon, 26 Jan 2009)*
1509 - Version 2 is a general rewrite of the code and therefore the
1510 differences are too many to list, see the design document for 2.0 in
1511 the ``doc/`` subdirectory for more details
1512 - In this beta version there is not yet a migration path from 1.2 (there
1513 will be one in the final 2.0 release)
1514 - A few significant changes are:
1516 - all commands are executed by a daemon (``ganeti-masterd``) and the
1517 various ``gnt-*`` commands are just front-ends to it
1518 - all the commands are entered into, and executed from a job queue,
1519 see the ``gnt-job(8)`` manpage
1520 - the RAPI daemon supports read-write operations, secured by basic
1521 HTTP authentication on top of HTTPS
1522 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1523 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1524 to migrate to DRBD 8 first while still running Ganeti 1.2)
1525 - DRBD devices are using statically allocated minor numbers, which
1526 will be assigned to existing instances during the migration process
1527 - there is support for both Xen PVM and Xen HVM instances running on
1529 - KVM virtualization is supported too
1530 - file-based storage has been implemented, which means that it is
1531 possible to run the cluster without LVM and DRBD storage, for
1532 example using a shared filesystem exported from shared storage (and
1533 still have live migration)
1539 *(Released Tue, 13 Jan 2009)*
1541 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1542 - Reuse the old instance mac address by default on instance import, if
1543 the instance name is the same.
1544 - Handle situations in which the node info rpc returns incomplete
1546 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1547 - Improved version of batcher:
1549 - state file support
1550 - instance mac address support
1551 - support for HVM clusters/instances
1553 - Add an option to show the number of cpu sockets and nodes in
1555 - Support OSes that handle more than one version of the OS api (but do
1556 not change the current API in any other way)
1557 - Fix ``gnt-node migrate``
1558 - ``gnt-debug`` man page
1559 - Fixes various more typos and small issues
1560 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1566 *(Released Wed, 24 Sep 2008)*
1568 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1569 type of disk exported to fully virtualized instances.
1570 - provide access to the serial console of HVM instances
1571 - instance auto_balance flag, set by default. If turned off it will
1572 avoid warnings on cluster verify if there is not enough memory to fail
1573 over an instance. in the future it will prevent automatically failing
1574 it over when we will support that.
1575 - batcher tool for instance creation, see ``tools/README.batcher``
1576 - ``gnt-instance reinstall --select-os`` to interactively select a new
1577 operating system when reinstalling an instance.
1578 - when changing the memory amount on instance modify a check has been
1579 added that the instance will be able to start. also warnings are
1580 emitted if the instance will not be able to fail over, if auto_balance
1582 - documentation fixes
1583 - sync fields between ``gnt-instance list/modify/add/import``
1584 - fix a race condition in drbd when the sync speed was set after giving
1585 the device a remote peer.
1591 *(Released Tue, 22 Jul 2008)*
1593 - note: the allowed size and number of tags per object were reduced
1594 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1595 - fixed twisted 8.x compatibility
1596 - fixed ``gnt-instance replace-disks`` with iallocator
1597 - add TCP keepalives on twisted connections to detect restarted nodes
1598 - disk increase support, see ``gnt-instance grow-disk``
1599 - implement bulk node/instance query for RAPI
1600 - add tags in node/instance listing (optional)
1601 - experimental migration (and live migration) support, read the man page
1602 for ``gnt-instance migrate``
1603 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1604 has some small improvements in handling its state file
1610 *(Released Fri, 13 Jun 2008)*
1612 - Experimental readonly, REST-based remote API implementation;
1613 automatically started on master node, TCP port 5080, if enabled by
1614 ``--enable-rapi`` parameter to configure script.
1615 - Instance allocator support. Add and import instance accept a
1616 ``--iallocator`` parameter, and call that instance allocator to decide
1617 which node to use for the instance. The iallocator document describes
1618 what's expected from an allocator script.
1619 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1620 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1621 node is lost there is still enough memory to fail over the instances
1623 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1624 ``gnt-cluster verify``, to check for site-specific compliance. All the
1625 hooks will run, and their output, if any, will be displayed. Any
1626 failing hook will make the verification return an error value.
1627 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1628 primary and secondary interfaces
1629 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1630 that is still declared as part of the cluster and has failed.
1631 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1632 specifying output fields, that just adds the chosen fields to the
1634 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1635 export from the filesystem.
1636 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1637 have been added. Using them you can enable/disable acpi and pae
1638 support, and specify a path for a cd image to be exported to the
1639 instance. These parameters as the name suggest only work on HVM
1641 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1642 PAE support will be set to the previously hardcoded values, but the
1643 (previously hardcoded) path to the CDROM ISO image will be unset and
1644 if required, needs to be set manually with ``gnt-instance modify``
1646 - The address to which an instance's VNC console is bound is now
1647 selectable per-instance, rather than being cluster wide. Of course
1648 this only applies to instances controlled via VNC, so currently just
1649 applies to HVM clusters.
1655 *(Released Mon, 18 Feb 2008)*
1657 - more tweaks to the disk activation code (especially helpful for DRBD)
1658 - change the default ``gnt-instance list`` output format, now there is
1659 one combined status field (see the manpage for the exact values this
1661 - some more fixes for the mac export to hooks change
1662 - make Ganeti not break with DRBD 8.2.x (which changed the version
1663 format in ``/proc/drbd``) (issue 24)
1664 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1665 template, allowing migration from DRBD0.7+MD to DRBD8
1671 *(Released Wed, 30 Jan 2008)*
1673 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1675 - add command aliases infrastructure and a few aliases
1676 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1677 man pages and the ``--help`` option of ``gnt-node
1678 list``/``gnt-instance list``
1679 - fix ``gnt-backup list`` with down nodes (issue 21)
1680 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1681 - fix the dist archive and add a check for including svn/git files in
1683 - some developer-related changes: improve the burnin and the QA suite,
1684 add an upload script for testing during development
1690 *(Released Wed, 16 Jan 2008)*
1692 - experimental HVM support, read the install document, section
1693 "Initializing the cluster"
1694 - allow for the PVM hypervisor per-instance kernel and initrd paths
1695 - add a new command ``gnt-cluster verify-disks`` which uses a new
1696 algorithm to improve the reconnection of the DRBD pairs if the device
1697 on the secondary node has gone away
1698 - make logical volume code auto-activate LVs at disk activation time
1699 - slightly improve the speed of activating disks
1700 - allow specification of the MAC address at instance creation time, and
1701 changing it later via ``gnt-instance modify``
1702 - fix handling of external commands that generate lots of output on
1704 - update documentation with regard to minimum version of DRBD8 supported
1710 *(Released Tue, 4 Dec 2007)*
1712 - Log the ``xm create`` output to the node daemon log on failure (to
1713 help diagnosing the error)
1714 - In debug mode, log all external commands output if failed to the logs
1715 - Change parsing of lvm commands to ignore stderr
1721 *(Released Wed, 28 Nov 2007)*
1723 - Another round of updates to the DRBD 8 code to deal with more failures
1724 in the replace secondary node operation
1725 - Some more logging of failures in disk operations (lvm, drbd)
1726 - A few documentation updates
1733 *(Released Tue, 13 Nov 2007)*
1735 - Change configuration file format from Python's Pickle to JSON.
1736 Upgrading is possible using the cfgupgrade utility.
1737 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1738 faster replace disks and is more stable (DRBD 8 has many improvements
1739 compared to DRBD 0.7)
1740 - Added command line tags support (see man pages for ``gnt-instance``,
1741 ``gnt-node``, ``gnt-cluster``)
1742 - Added instance rename support
1743 - Added multi-instance startup/shutdown
1744 - Added cluster rename support
1745 - Added ``gnt-node evacuate`` to simplify some node operations
1746 - Added instance reboot operation that can speedup reboot as compared to
1748 - Soften the requirement that hostnames are in FQDN format
1749 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1751 - Removed dependency on debian's patched fping that uses the
1752 non-standard ``-S`` option
1753 - Now the OS definitions are searched for in multiple, configurable
1754 paths (easier for distros to package)
1755 - Some changes to the hooks infrastructure (especially the new
1756 post-configuration update hook)
1757 - Other small bugfixes
1759 .. vim: set textwidth=72 syntax=rst :