7 *(Released Mon, 22 Aug 2011)*
9 Incompatible/important changes and bugfixes
10 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
12 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
13 resource's ``ip_check`` parameter changed from ``True`` to ``False``
14 to match the underlying LUXI interface.
15 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
16 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
17 not maintain backwards-compatibility as the underlying operation
18 changed in an incompatible way. The RAPI client can talk to old
19 servers, but it needs to be told so as the return value changed.
20 - When creating file-based instances via RAPI, the ``file_driver``
21 parameter no longer defaults to ``loop`` and must be specified.
22 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
24 - Support for the undocumented and deprecated RAPI instance creation
25 request format version 0 has been dropped. Use version 1, supported
26 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
27 - Pyparsing 1.4.6 or above is required, see :doc:`installation
28 documentation <install>`.
29 - The "cluster-verify" hooks are now executed per group by the
30 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
31 if you just run ``gnt-cluster verify``, which generates one opcode per
33 - The environment as passed to the OS scripts is cleared, and thus no
34 environment variables defined in the node daemon's environment will be
35 inherited by the scripts.
36 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
38 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
39 support operations involving multiple node groups.
40 - Offline nodes are ignored when failing over an instance.
45 - The ganeti-htools project has been merged into the ganeti-core source
46 tree and will be built as part of Ganeti (see :doc:`install-quick`).
47 - Implemented support for :doc:`shared storage <design-shared-storage>`.
48 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
49 GPT-style partition tables (requires `parted
50 <http://www.gnu.org/s/parted/>`_).
51 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
52 - Allowed adding tags on instance creation.
53 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
54 - Allow instances to be started in a paused state, enabling the user to
55 see the complete console output on boot using the console.
56 - Added new hypervisor flag to control default reboot behaviour
57 (``reboot_behavior``).
58 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
59 - Improved out-of-band management support:
61 - Added ``gnt-node health`` command reporting the health status of
63 - Added ``gnt-node power`` command to manage power status of nodes.
64 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
66 - Instance migration can fall back to failover if instance is not
68 - Filters can be used when listing nodes, instances, groups and locks;
69 see *ganeti(7)* manpage.
70 - Added post-execution status as variables to :doc:`hooks <hooks>`
72 - Instance tags are exported/imported together with the instance.
73 - When given an explicit job ID, ``gnt-job info`` will work for archived
75 - Jobs can define dependencies on other jobs (not yet supported via
76 RAPI or command line, but used by internal commands and usable via
79 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
82 - Instance failover is now available as a RAPI resource
83 (``/2/instances/[instance_name]/failover``).
84 - ``gnt-instance info`` defaults to static information if primary node
86 - Opcodes have a new ``comment`` attribute.
87 - Added basic SPICE support to KVM hypervisor.
89 Node group improvements
90 ~~~~~~~~~~~~~~~~~~~~~~~
92 - ``gnt-cluster verify`` has been modified to check groups separately,
93 thereby improving performance.
94 - Node group support has been added to ``gnt-cluster verify-disks``,
95 which now operates per node group.
96 - Watcher has been changed to work better with node groups.
98 - One process and state file per node group.
99 - Slow watcher in one group doesn't block other group's watcher.
101 - Added new command, ``gnt-group evacuate``, to move all instances in a
102 node group to other groups.
103 - Added ``gnt-instance change-group`` to move an instance to another
105 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
106 per-group operations.
107 - Node groups can be tagged.
108 - Some operations switch from an exclusive to a shared lock as soon as
110 - Instance's primary and secondary nodes' groups are now available as
111 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
112 and ``snodes.group.uuid``).
117 - Numerous updates to documentation and manpages.
119 - :doc:`RAPI <rapi>` documentation now has detailed parameter
122 - A lockset's internal lock is now also visible in lock monitor.
123 - Log messages from job queue workers now contain information about the
124 opcode they're processing.
125 - ``gnt-instance console`` no longer requires the instance lock.
126 - A short delay when waiting for job changes reduces the number of LUXI
127 requests significantly.
128 - DRBD metadata volumes are overwritten with zeros during disk creation.
134 *(Released Fri, 12 Aug 2011)*
136 This was the first beta release of the 2.5 series.
142 *(Released Fri, 5 Aug 2011)*
144 Many bug-fixes and a few small features:
146 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
147 issues when you tried to add an instance with two MAC addresses in one
149 - KVM: fixed per-instance stored UID value
150 - KVM: configure bridged NICs at migration start
151 - KVM: Fix a bug where instance will not start with never KVM versions
153 - Added OS search path to ``gnt-cluster info``
154 - Fixed an issue with ``file_storage_dir`` where you were forced to
155 provide an absolute path, but the documentation states it is a
156 relative path, the documentation was right
157 - Added a new parameter to instance stop/start called ``--no-remember``
158 that will make the state change to not be remembered
159 - Implemented ``no_remember`` at RAPI level
160 - Improved the documentation
161 - Node evacuation: don't call IAllocator if node is already empty
162 - Fixed bug in DRBD8 replace disks on current nodes
163 - Fixed bug in recreate-disks for DRBD instances
164 - Moved assertion checking locks in ``gnt-instance replace-disks``
165 causing it to abort with not owning the right locks for some situation
166 - Job queue: Fixed potential race condition when cancelling queued jobs
167 - Fixed off-by-one bug in job serial generation
168 - ``gnt-node volumes``: Fix instance names
169 - Fixed aliases in bash completion
170 - Fixed a bug in reopening log files after beeing sent a SIGHUP
171 - Added a flag to burnin to allow specifying VCPU count
172 - Bugfixes to non-root Ganeti configuration
178 *(Released Thu, 12 May 2011)*
180 Many bug-fixes and a few new small features:
182 - Fixed a bug related to log opening failures
183 - Fixed a bug in instance listing with orphan instances
184 - Fixed a bug which prevented resetting the cluster-level node parameter
185 ``oob_program`` to the default
186 - Many fixes related to the ``cluster-merge`` tool
187 - Fixed a race condition in the lock monitor, which caused failures
188 during (at least) creation of many instances in parallel
189 - Improved output for gnt-job info
190 - Removed the quiet flag on some ssh calls which prevented debugging
192 - Improved the N+1 failure messages in cluster verify by actually
193 showing the memory values (needed and available)
194 - Increased lock attempt timeouts so that when executing long operations
195 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
196 too early and thus prevent the use of the 'fair' mechanism
197 - Changed instance query data (``gnt-instance info``) to not acquire
198 locks unless needed, thus allowing its use on locked instance if only
199 static information is asked for
200 - Improved behaviour with filesystems that do not support rename on an
202 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
203 kept locks on all nodes during the wipe, which is unneeded
204 - Fixed ``gnt-watcher`` handling of errors during hooks execution
205 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
206 10GiB) which caused the wipe to fail right at the end in some cases
207 - Fixed master IP activation when doing master failover with no-voting
208 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
209 the master node itself
210 - Fixed potential data-loss in under disk full conditions, where Ganeti
211 wouldn't check correctly the return code and would consider
212 partially-written files 'correct'
213 - Fixed bug related to multiple VGs and DRBD disk replacing
214 - Added new disk parameter ``metavg`` that allows placement of the meta
215 device for DRBD in a different volume group
216 - Fixed error handling in the node daemon when the system libc doesn't
217 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
218 - Fixed lock release during replace-disks, which kept cluster-wide locks
219 when doing disk replaces with an iallocator script
220 - Added check for missing bridges in cluster verify
221 - Handle EPIPE errors while writing to the terminal better, so that
222 piping the output to e.g. ``less`` doesn't cause a backtrace
223 - Fixed rare case where a ^C during Luxi calls could have been
224 interpreted as server errors, instead of simply terminating
225 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
227 - Added a few more parameters to the KVM hypervisor, allowing a second
228 CDROM, custom disk type for CDROMs and a floppy image
229 - Removed redundant message in instance rename when the name is given
231 - Added option to ``gnt-instance recreate-disks`` to allow creating the
232 disks on new nodes, allowing recreation when the original instance
233 nodes are completely gone
234 - Added option when converting disk templates to DRBD to skip waiting
235 for the resync, in order to make the instance available sooner
236 - Added two new variables to the OS scripts environment (containing the
238 - Made the root_path and optional parameter for the xen-pvm hypervisor,
239 to allow use of ``pvgrub`` as bootloader
240 - Changed the instance memory modifications to only check out-of-memory
241 conditions on memory increases, and turned the secondary node warnings
242 into errors (they can still be overridden via ``--force``)
243 - Fixed the handling of a corner case when the Python installation gets
244 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
245 to execute a command that doesn't exist
246 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
247 node of the instance returned failures during instance shutdown; this
248 adds the option ``--ignore-consistency`` to gnt-instance move
250 And as usual, various improvements to the error messages, documentation
256 *(Released Wed, 09 Mar 2011)*
258 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
259 the RAPI users file if run twice (even with ``--dry-run``).
261 The release fixes that bug (nothing else changed).
267 *(Released Mon, 07 Mar 2011)*
269 Final 2.4.0 release. Just a few small fixes:
271 - Fixed RAPI node evacuate
272 - Fixed the kvm-ifup script
273 - Fixed internal error handling for special job cases
274 - Updated man page to specify the escaping feature for options
280 *(Released Mon, 28 Feb 2011)*
282 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
283 that this feature wiped the disks of the wrong instance, leading to loss
288 - Fixed title of query field containing instance name
289 - Expanded the glossary in the documentation
290 - Fixed one unittest (internal issue)
296 *(Released Mon, 21 Feb 2011)*
298 A number of bug fixes plus just a couple functionality changes.
300 On the user-visible side, the ``gnt-* list`` command output has changed
301 with respect to "special" field states. The current rc1 style of display
302 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
303 the default output mode special fields states are displayed as follows:
305 - Offline resource: ``*``
306 - Unavailable/not applicable: ``-``
307 - Data missing (RPC failure): ``?``
308 - Unknown field: ``??``
310 Another user-visible change is the addition of ``--force-join`` to
315 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
316 - Fixed regression in RAPI/instance reinstall where all parameters were
317 required (instead of optional)
318 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
319 - Fixed iallocator usage (offline nodes were not considered offline)
320 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
321 - Fixed hypervisor and OS parameter validation with respect to
323 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
325 - Fixed ``tools/listrunner`` with respect to agent-based usage
331 *(Released Fri, 4 Feb 2011)*
333 Many changes and fixes since the beta1 release. While there were some
334 internal changes, the code has been mostly stabilised for the RC
337 Note: the dumb allocator was removed in this release, as it was not kept
338 up-to-date with the IAllocator protocol changes. It is recommended to
339 use the ``hail`` command from the ganeti-htools package.
341 Note: the 2.4 and up versions of Ganeti are not compatible with the
342 0.2.x branch of ganeti-htools. You need to upgrade to
343 ganeti-htools-0.3.0 (or later).
345 Regressions fixed from 2.3
346 ~~~~~~~~~~~~~~~~~~~~~~~~~~
348 - Fixed the ``gnt-cluster verify-disks`` command
349 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
351 - Fixed disk adoption breakage
352 - Fixed wrong headers in instance listing for field aliases
357 - Fixed corner case in KVM handling of NICs
358 - Fixed many cases of wrong handling of non-vm_capable nodes
359 - Fixed a bug where a missing instance symlink was not possible to
360 recreate with any ``gnt-*`` command (now ``gnt-instance
361 activate-disks`` does it)
362 - Fixed the volume group name as reported by ``gnt-cluster
364 - Increased timeouts for the import-export code, hopefully leading to
365 fewer aborts due network or instance timeouts
366 - Fixed bug in ``gnt-node list-storage``
367 - Fixed bug where not all daemons were started on cluster
368 initialisation, but only at the first watcher run
369 - Fixed many bugs in the OOB implementation
370 - Fixed watcher behaviour in presence of instances with offline
372 - Fixed instance list output for instances running on the wrong node
373 - a few fixes to the cluster-merge tool, but it still cannot merge
374 multi-node groups (currently it is not recommended to use this tool)
380 - Improved network configuration for the KVM hypervisor
381 - Added e1000 as a supported NIC for Xen-HVM
382 - Improved the lvmstrap tool to also be able to use partitions, as
383 opposed to full disks
384 - Improved speed of disk wiping (the cluster parameter
385 ``prealloc_wipe_disks``, so that it has a low impact on the total time
386 of instance creations
387 - Added documentation for the OS parameters
388 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
389 hypervisor is not responding
390 - Added display of blacklisted and hidden OS information in
392 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
393 NIC and node parameters, which might create problems with currently
394 invalid (but undetected) configuration files, but prevents validation
395 failures when unrelated parameters are modified
396 - Changed cluster initialisation to wait for the master daemon to become
398 - Expanded the RAPI interface:
400 - Added config redistribution resource
401 - Added activation/deactivation of instance disks
402 - Added export of console information
404 - Implemented log file reopening on SIGHUP, which allows using
405 logrotate(8) for the Ganeti log files
406 - Added a basic OOB helper script as an example
412 *(Released Fri, 14 Jan 2011)*
417 - Fixed timezone issues when formatting timestamps
418 - Added support for node groups, available via ``gnt-group`` and other
420 - Added out-of-band framework and management, see :doc:`design
421 document <design-oob>`
422 - Removed support for roman numbers from ``gnt-node list`` and
423 ``gnt-instance list``.
424 - Allowed modification of master network interface via ``gnt-cluster
425 modify --master-netdev``
426 - Accept offline secondaries while shutting down instance disks
427 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
428 - Added support for multiple LVM volume groups
429 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
431 - Added commands to list available fields:
433 - ``gnt-node list-fields``
434 - ``gnt-group list-fields``
435 - ``gnt-instance list-fields``
437 - Updated documentation and man pages
442 - Moved ``rapi_users`` file into separate directory, now named
443 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
445 - Added new tool for running commands on many machines,
446 ``tools/ganeti-listrunner``
447 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
448 improving the ``gnt-instance console`` output
449 - Allowed customisation of disk index separator at ``configure`` time
450 - Export node group allocation policy to :doc:`iallocator <iallocator>`
451 - Added support for non-partitioned md disks in ``lvmstrap``
452 - Added script to gracefully power off KVM instances
453 - Split ``utils`` module into smaller parts
454 - Changed query operations to return more detailed information, e.g.
455 whether an information is unavailable due to an offline node. To use
456 this new functionality, the LUXI call ``Query`` must be used. Field
457 information is now stored by the master daemon and can be retrieved
458 using ``QueryFields``. Instances, nodes and groups can also be queried
459 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
460 exposed via RAPI). The following commands make use of this
461 infrastructure change:
464 - ``gnt-group list-fields``
466 - ``gnt-node list-fields``
467 - ``gnt-instance list``
468 - ``gnt-instance list-fields``
469 - ``gnt-debug locks``
474 - New RAPI resources (see :doc:`rapi`):
478 - ``/2/groups/[group_name]``
479 - ``/2/groups/[group_name]/assign-nodes``
480 - ``/2/groups/[group_name]/modify``
481 - ``/2/groups/[group_name]/rename``
482 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
486 - Implemented ``no_install`` for instance creation
487 - Implemented OS parameters for instance reinstallation, allowing
488 use of special settings on reinstallation (e.g. for preserving data)
493 - Added IPv6 support in import/export
494 - Pause DRBD synchronization while wiping disks on instance creation
495 - Updated unittests and QA scripts
496 - Improved network parameters passed to KVM
497 - Converted man pages from docbook to reStructuredText
503 *(Released Mon, 20 Dec 2010)*
505 Released version 2.3.1~rc1 without any changes.
511 *(Released Wed, 1 Dec 2010)*
513 - impexpd: Disable OpenSSL compression in socat if possible (backport
514 from master, commit e90739d625b, see :doc:`installation guide
515 <install-quick>` for details)
516 - Changed unittest coverage report to exclude test scripts
517 - Added script to check version format
523 *(Released Wed, 1 Dec 2010)*
525 Released version 2.3.0~rc1 without any changes.
531 *(Released Fri, 19 Nov 2010)*
533 A number of bugfixes and documentation updates:
535 - Update ganeti-os-interface documentation
536 - Fixed a bug related to duplicate MACs or similar items which should be
538 - Fix breakage in OS state modify
539 - Reinstall instance: disallow offline secondaries (fixes bug related to
540 OS changing but reinstall failing)
541 - plus all the other fixes between 2.2.1 and 2.2.2
547 *(Released Tue, 2 Nov 2010)*
549 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
550 - Fixed master failover race with watcher
551 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
553 - Accept previously stopped instance for export with instance removal
554 - Simplify and extend the environment variables for instance OS scripts
555 - Added new node flags, ``master_capable`` and ``vm_capable``
556 - Added optional instance disk wiping prior during allocation. This is a
557 cluster-wide option and can be set/modified using
558 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
559 - Added IPv6 support, see :doc:`design document <design-2.3>` and
561 - Added a new watcher option (``--ignore-pause``)
562 - Added option to ignore offline node on instance start/stop
563 (``--ignore-offline``)
564 - Allow overriding OS parameters with ``gnt-instance reinstall``
565 - Added ability to change node's secondary IP address using ``gnt-node
567 - Implemented privilege separation for all daemons except
568 ``ganeti-noded``, see ``configure`` options
569 - Complain if an instance's disk is marked faulty in ``gnt-cluster
571 - Implemented job priorities (see ``ganeti(7)`` manpage)
572 - Ignore failures while shutting down instances during failover from
574 - Exit daemon's bootstrap process only once daemon is ready
575 - Export more information via ``LUInstanceQuery``/remote API
576 - Improved documentation, QA and unittests
577 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
578 restart if the file was created or changed
579 - Added LUXI protocol version sent with each request and response,
580 allowing detection of server/client mismatches
581 - Moved the Python scripts among gnt-* and ganeti-* into modules
582 - Moved all code related to setting up SSH to an external script,
584 - Infrastructure changes for node group support in future versions
590 *(Released Fri, 19 Nov 2010)*
592 A few small bugs fixed, and some improvements to the build system:
594 - Fix documentation regarding conversion to drbd
595 - Fix validation of parameters in cluster modify (``gnt-cluster modify
597 - Fix error handling in node modify with multiple changes
598 - Allow remote imports without checked names
604 *(Released Tue, 19 Oct 2010)*
606 - Disable SSL session ID cache in RPC client
612 *(Released Thu, 14 Oct 2010)*
614 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
615 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
617 - Fix problems with interaction between Curl and Python's HTTP server,
618 resulting in increased speed in many RPC calls
619 - Improve our release script to prevent breakage with older aclocal and
626 *(Released Thu, 7 Oct 2010)*
628 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
629 value retrieved from master
630 - Added support for blacklisted or hidden OS definitions
631 - Added simple lock monitor (accessible via (``gnt-debug locks``)
632 - Added support for -mem-path in KVM hypervisor abstraction layer
633 - Allow overriding instance parameters in tool for inter-cluster
634 instance moves (``tools/move-instance``)
635 - Improved opcode summaries (e.g. in ``gnt-job list``)
636 - Improve consistency of OS listing by sorting it
637 - Documentation updates
643 *(Released Fri, 8 Oct 2010)*
645 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
651 *(Released Mon, 4 Oct 2010)*
653 - Fixed regression in ``gnt-instance rename``
659 *(Released Wed, 22 Sep 2010)*
661 - Fixed OS_VARIANT variable for OS scripts
662 - Fixed cluster tag operations via RAPI
663 - Made ``setup-ssh`` exit with non-zero code if an error occurred
664 - Disabled RAPI CA checks in watcher
670 *(Released Mon, 23 Aug 2010)*
672 - Support DRBD versions of the format "a.b.c.d"
674 - Re-introduce support for usage from multiple threads in RAPI client
675 - Instance renames and modify via RAPI
676 - Work around race condition between processing and archival in job
678 - Mark opcodes following failed one as failed, too
679 - Job field ``lock_status`` was removed due to difficulties making it
680 work with the changed job queue in Ganeti 2.2; a better way to monitor
681 locks is expected for a later 2.2.x release
682 - Fixed dry-run behaviour with many commands
683 - Support ``ssh-agent`` again when adding nodes
684 - Many additional bugfixes
690 *(Released Fri, 30 Jul 2010)*
692 Important change: the internal RPC mechanism between Ganeti nodes has
693 changed from using a home-grown http library (based on the Python base
694 libraries) to use the PycURL library. This requires that PycURL is
695 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
696 against GnuTLS by default. cURL's support for GnuTLS had known issues
697 before cURL 7.21.0 and we recommend using the latest cURL release or
698 linking against OpenSSL. Most other distributions already link PycURL
699 and cURL against OpenSSL. The command::
701 python -c 'import pycurl; print pycurl.version'
703 can be used to determine the libraries PycURL and cURL are linked
706 Other significant changes:
708 - Rewrote much of the internals of the job queue, in order to achieve
709 better parallelism; this decouples job query operations from the job
710 processing, and it should allow much nicer behaviour of the master
711 daemon under load, and it also has uncovered some long-standing bugs
712 related to the job serialisation (now fixed)
713 - Added a default iallocator setting to the cluster parameters,
714 eliminating the need to always pass nodes or an iallocator for
715 operations that require selection of new node(s)
716 - Added experimental support for the LXC virtualization method
717 - Added support for OS parameters, which allows the installation of
718 instances to pass parameter to OS scripts in order to customise the
720 - Added a hypervisor parameter controlling the migration type (live or
721 non-live), since hypervisors have various levels of reliability; this
722 has renamed the 'live' parameter to 'mode'
723 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
724 logical volumes, meaning that cluster verify will ignore them and not
725 flag their presence as errors
726 - The watcher will now reset the error count for failed instances after
727 8 hours, thus allowing self-healing if the problem that caused the
728 instances to be down/fail to start has cleared in the meantime
729 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
730 check for, and warn, if the drbd module parameter ``usermode_helper``
731 is not consistent with the cluster-wide setting; this is needed to
732 make diagnose easier of failed drbd creations
733 - Started adding base IPv6 support, but this is not yet
734 enabled/available for use
735 - Rename operations (cluster, instance) will now return the new name,
736 which is especially useful if a short name was passed in
737 - Added support for instance migration in RAPI
738 - Added a tool to pre-configure nodes for the SSH setup, before joining
739 them to the cluster; this will allow in the future a simplified model
740 for node joining (but not yet fully enabled in 2.2); this needs the
741 paramiko python library
742 - Fixed handling of name-resolving errors
743 - Fixed consistency of job results on the error path
744 - Fixed master-failover race condition when executed multiple times in
746 - Fixed many bugs related to the job queue (mostly introduced during the
747 2.2 development cycle, so not all are impacting 2.1)
748 - Fixed instance migration with missing disk symlinks
749 - Fixed handling of unknown jobs in ``gnt-job archive``
750 - And many other small fixes/improvements
754 - Enhanced both the unittest and the QA coverage
755 - Switched the opcode validation to a generic model, and extended the
756 validation to all opcode parameters
757 - Changed more parts of the code that write shell scripts to use the
759 - Switched the master daemon to use the asyncore library for the Luxi
766 *(Released Thu, 17 Jun 2010)*
768 - Added tool (``move-instance``) and infrastructure to move instances
769 between separate clusters (see :doc:`separate documentation
770 <move-instance>` and :doc:`design document <design-2.2>`)
771 - Added per-request RPC timeout
772 - RAPI now requires a Content-Type header for requests with a body (e.g.
773 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
774 :rfc:`2616` (HTTP/1.1), section 7.2.1)
775 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
777 - Implemented initial support for running Ganeti daemons as separate
778 users, see configure-time flags ``--with-user-prefix`` and
779 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
781 - Instances can be removed after export (``gnt-backup export
783 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
784 key (instead of 1024 bit)
785 - Added new cluster configuration file for cluster domain secret
786 - Import/export now use SSL instead of SSH
787 - Added support for showing estimated time when exporting an instance,
788 see the ``ganeti-os-interface(7)`` manpage and look for
795 *(Released Tue, 16 Nov 2010)*
797 Some more bugfixes. Unless critical bugs occur, this will be the last
800 - Fix case of MAC special-values
801 - Fix mac checker regex
802 - backend: Fix typo causing "out of range" error
803 - Add missing --units in gnt-instance list man page
809 *(Released Tue, 24 Aug 2010)*
812 - Don't ignore secondary node silently on non-mirrored disk templates
814 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
815 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
816 - Properly document the usb_mouse parameter
817 - Fix path in ganeti-rapi(8) (issue 116)
818 - Adjust error message when the ganeti user's .ssh directory is
820 - Add same-node-check when changing the disk template to drbd
826 *(Released Fri, 16 Jul 2010)*
829 - Add an option to only select some reboot types during qa/burnin.
830 (on some hypervisors consequent reboots are not supported)
831 - Fix infrequent race condition in master failover. Sometimes the old
832 master ip address would be still detected as up for a short time
833 after it was removed, causing failover to fail.
834 - Decrease mlockall warnings when the ctypes module is missing. On
835 Python 2.4 we support running even if no ctypes module is installed,
836 but we were too verbose about this issue.
837 - Fix building on old distributions, on which man doesn't have a
839 - Fix RAPI not to ignore the MAC address on instance creation
840 - Implement the old instance creation format in the RAPI client.
846 *(Released Thu, 01 Jul 2010)*
848 A small bugfix release:
849 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
850 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
851 a non-existing option
852 - Fix instance create: the --force-variant option was ignored
853 - Improve pylint 0.21 compatibility and warnings with Python 2.6
854 - Fix modify node storage with non-FQDN arguments
855 - Fix RAPI client to authenticate under Python 2.6 when used
856 for more than 5 requests needing authentication
857 - Fix gnt-instance modify -t (storage) giving a wrong error message
858 when converting a non-shutdown drbd instance to plain
864 *(Released Fri, 18 Jun 2010)*
866 A small bugfix release:
868 - Fix live migration of KVM instances started with older Ganeti
869 versions which had fewer hypervisor parameters
870 - Fix gnt-instance grow-disk on down instances
871 - Fix an error-reporting bug during instance migration
872 - Better checking of the ``--net`` and ``--disk`` values, to avoid
873 silently ignoring broken ones
874 - Fix an RPC error reporting bug affecting, for example, RAPI client
876 - Fix bug triggered by different API version os-es on different nodes
877 - Fix a bug in instance startup with custom hvparams: OS level
878 parameters would fail to be applied.
879 - Fix the RAPI client under Python 2.6 (but more work is needed to
880 make it work completely well with OpenSSL)
881 - Fix handling of errors when resolving names from DNS
887 *(Released Thu, 3 Jun 2010)*
889 A medium sized development cycle. Some new features, and some
890 fixes/small improvements/cleanups.
895 The node deamon now tries to mlock itself into memory, unless the
896 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
897 its logs, and falls back to console logging. This allows emergency
898 features such as ``gnt-node powercycle`` to work even in the event of a
899 broken node disk (tested offlining the disk hosting the node's
900 filesystem and dropping its memory caches; don't try this at home)
902 KVM: add vhost-net acceleration support. It can be tested with a new
903 enough version of the kernel and of qemu-kvm.
905 KVM: Add instance chrooting feature. If you use privilege dropping for
906 your VMs you can also now force them to chroot to an empty directory,
907 before starting the emulated guest.
909 KVM: Add maximum migration bandwith and maximum downtime tweaking
910 support (requires a new-enough version of qemu-kvm).
912 Cluster verify will now warn if the master node doesn't have the master
915 Add a new (incompatible) instance creation request format to RAPI which
916 supports all parameters (previously only a subset was supported, and it
917 wasn't possible to extend the old format to accomodate all the new
918 features. The old format is still supported, and a client can check for
919 this feature, before using it, by checking for its presence in the
920 ``features`` RAPI resource.
922 Now with ancient latin support. Try it passing the ``--roman`` option to
923 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
924 (requires the python-roman module to be installed, in order to work).
929 As usual many internal code refactorings, documentation updates, and
932 - Lots of improvements and cleanups to the experimental Remote API
933 (RAPI) client library.
934 - A new unit test suite for the core daemon libraries.
935 - A fix to creating missing directories makes sure the umask is not
936 applied anymore. This enforces the same directory permissions
938 - Better handling terminating daemons with ctrl+c (used when running
939 them in debugging mode).
940 - Fix a race condition in live migrating a KVM instance, when stat()
941 on the old proc status file returned EINVAL, which is an unexpected
943 - Fixed manpage checking with newer man and utf-8 charachters. But now
944 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
950 *(Released Fri, 7 May 2010)*
952 Fix a bug which prevented untagged KVM instances from starting.
958 *(Released Fri, 7 May 2010)*
960 Another release with a long development cycle, during which many
961 different features were added.
966 The KVM hypervisor now can run the individual instances as non-root, to
967 reduce the impact of a VM being hijacked due to bugs in the
968 hypervisor. It is possible to run all instances as a single (non-root)
969 user, to manually specify a user for each instance, or to dynamically
970 allocate a user out of a cluster-wide pool to each instance, with the
971 guarantee that no two instances will run under the same user ID on any
974 An experimental RAPI client library, that can be used standalone
975 (without the other Ganeti libraries), is provided in the source tree as
976 ``lib/rapi/client.py``. Note this client might change its interface in
977 the future, as we iterate on its capabilities.
979 A new command, ``gnt-cluster renew-crypto`` has been added to easily
980 replace the cluster's certificates and crypto keys. This might help in
981 case they have been compromised, or have simply expired.
983 A new disk option for instance creation has been added that allows one
984 to "adopt" currently existing logical volumes, with data
985 preservation. This should allow easier migration to Ganeti from
986 unmanaged (or managed via other software) instances.
988 Another disk improvement is the possibility to convert between redundant
989 (DRBD) and plain (LVM) disk configuration for an instance. This should
990 allow better scalability (starting with one node and growing the
991 cluster, or shrinking a two-node cluster to one node).
993 A new feature that could help with automated node failovers has been
994 implemented: if a node sees itself as offline (by querying the master
995 candidates), it will try to shutdown (hard) all instances and any active
996 DRBD devices. This reduces the risk of duplicate instances if an
997 external script automatically failovers the instances on such nodes. To
998 enable this, the cluster parameter ``maintain_node_health`` should be
999 enabled; in the future this option (per the name) will enable other
1000 automatic maintenance features.
1002 Instance export/import now will reuse the original instance
1003 specifications for all parameters; that means exporting an instance,
1004 deleting it and the importing it back should give an almost identical
1005 instance. Note that the default import behaviour has changed from
1006 before, where it created only one NIC; now it recreates the original
1009 Cluster verify has added a few new checks: SSL certificates validity,
1010 /etc/hosts consistency across the cluster, etc.
1015 As usual, many internal changes were done, documentation fixes,
1018 - Fixed cluster initialization with disabled cluster storage (regression
1019 introduced in 2.1.1)
1020 - File-based storage supports growing the disks
1021 - Fixed behaviour of node role changes
1022 - Fixed cluster verify for some corner cases, plus a general rewrite of
1023 cluster verify to allow future extension with more checks
1024 - Fixed log spamming by watcher and node daemon (regression introduced
1026 - Fixed possible validation issues when changing the list of enabled
1028 - Fixed cleanup of /etc/hosts during node removal
1029 - Fixed RAPI response for invalid methods
1030 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1031 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1032 from ide disks, etc.)
1033 - Allow OS changes without re-installation (to record a changed OS
1034 outside of Ganeti, or to allow OS renames)
1035 - Allow instance creation without OS installation (useful for example if
1036 the OS will be installed manually, or restored from a backup not in
1038 - Implemented option to make cluster ``copyfile`` use the replication
1040 - Added list of enabled hypervisors to ssconf (possibly useful for
1042 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1044 - A partial form of node re-IP is possible via node readd, which now
1045 allows changed node primary IP
1046 - Command line utilities now show an informational message if the job is
1048 - The logs of the master daemon now show the PID/UID/GID of the
1055 *(Released Fri, 12 Mar 2010)*
1057 During the 2.1.0 long release candidate cycle, a lot of improvements and
1058 changes have accumulated with were released later as 2.1.1.
1063 The node evacuate command (``gnt-node evacuate``) was significantly
1064 rewritten, and as such the IAllocator protocol was changed - a new
1065 request type has been added. This unfortunate change during a stable
1066 series is designed to improve performance of node evacuations; on
1067 clusters with more than about five nodes and which are well-balanced,
1068 evacuation should proceed in parallel for all instances of the node
1069 being evacuated. As such, any existing IAllocator scripts need to be
1070 updated, otherwise the above command will fail due to the unknown
1071 request. The provided "dumb" allocator has not been updated; but the
1072 ganeti-htools package supports the new protocol since version 0.2.4.
1074 Another important change is increased validation of node and instance
1075 names. This might create problems in special cases, if invalid host
1076 names are being used.
1078 Also, a new layer of hypervisor parameters has been added, that sits at
1079 OS level between the cluster defaults and the instance ones. This allows
1080 customisation of virtualization parameters depending on the installed
1081 OS. For example instances with OS 'X' may have a different KVM kernel
1082 (or any other parameter) than the cluster defaults. This is intended to
1083 help managing a multiple OSes on the same cluster, without manual
1084 modification of each instance's parameters.
1086 A tool for merging clusters, ``cluster-merge``, has been added in the
1087 tools sub-directory.
1092 - Improved the int/float conversions that should make the code more
1093 robust in face of errors from the node daemons
1094 - Fixed the remove node code in case of internal configuration errors
1095 - Fixed the node daemon behaviour in face of inconsistent queue
1096 directory (e.g. read-only file-system where we can't open the files
1098 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1099 now it either aborts cleanly or, if given the new "auto_promote"
1100 parameter, will automatically promote other nodes as needed
1101 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1102 completely prevent Ganeti from working
1103 - Fixed bug for instance export when not all disks were successfully
1105 - Fixed behaviour of node add when the new node is slow in starting up
1107 - Fixed handling of signals in the LUXI client, which should improve
1108 behaviour of command-line scripts
1109 - Added checks for invalid node/instance names in the configuration (now
1110 flagged during cluster verify)
1111 - Fixed watcher behaviour for disk activation errors
1112 - Fixed two potentially endless loops in http library, which led to the
1113 RAPI daemon hanging and consuming 100% CPU in some cases
1114 - Fixed bug in RAPI daemon related to hashed passwords
1115 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1117 - Enhanced compatibility with non-Debian OSes, but not using absolute
1118 path in some commands and allowing customisation of the ssh
1119 configuration directory
1120 - Fixed possible future issue with new Python versions by abiding to the
1121 proper use of ``__slots__`` attribute on classes
1122 - Added checks that should prevent directory traversal attacks
1123 - Many documentation fixes based on feedback from users
1128 - Added an "early_release" more for instance replace disks and node
1129 evacuate, where we release locks earlier and thus allow higher
1130 parallelism within the cluster
1131 - Added watcher hooks, intended to allow the watcher to restart other
1132 daemons (e.g. from the ganeti-nbma project), but they can be used of
1133 course for any other purpose
1134 - Added a compile-time disable for DRBD barriers, to increase
1135 performance if the administrator trusts the power supply or the
1136 storage system to not lose writes
1137 - Added the option of using syslog for logging instead of, or in
1138 addition to, Ganeti's own log files
1139 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1140 can indeed boot from a paravirtual NIC
1141 - Added a generic debug level for many operations; while this is not
1142 used widely yet, it allows one to pass the debug value all the way to
1144 - Enhanced the hooks environment for instance moves (failovers,
1145 migrations) where the primary/secondary nodes changed during the
1146 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1147 - Enhanced data validations for many user-supplied values; one important
1148 item is the restrictions imposed on instance and node names, which
1149 might reject some (invalid) host names
1150 - Add a configure-time option to disable file-based storage, if it's not
1151 needed; this allows greater security separation between the master
1152 node and the other nodes from the point of view of the inter-node RPC
1154 - Added user notification in interactive tools if job is waiting in the
1155 job queue or trying to acquire locks
1156 - Added log messages when a job is waiting for locks
1157 - Added filtering by node tags in instance operations which admit
1158 multiple instances (start, stop, reboot, reinstall)
1159 - Added a new tool for cluster mergers, ``cluster-merge``
1160 - Parameters from command line which are of the form ``a=b,c=d`` can now
1161 use backslash escapes to pass in values which contain commas,
1162 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1164 - For KVM, the instance name is the first parameter passed to KVM, so
1165 that it's more visible in the process list
1171 *(Released Tue, 2 Mar 2010)*
1173 Ganeti 2.1 brings many improvements with it. Major changes:
1175 - Added infrastructure to ease automated disk repairs
1176 - Added new daemon to export configuration data in a cheaper way than
1177 using the remote API
1178 - Instance NICs can now be routed instead of being associated with a
1180 - Improved job locking logic to reduce impact of jobs acquiring multiple
1181 locks waiting for other long-running jobs
1183 In-depth implementation details can be found in the Ganeti 2.1 design
1189 - Added chroot hypervisor
1190 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1192 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1193 ``bootloader_path`` and ``bootloader_args``)
1194 - Added the ``use_localtime`` option for the xen-hvm and kvm
1195 hypervisors, and the default value for this has changed to false (in
1196 2.0 xen-hvm always enabled it)
1197 - Added luxi call to submit multiple jobs in one go
1198 - Added cluster initialization option to not modify ``/etc/hosts``
1200 - Added network interface parameters
1201 - Added dry run mode to some LUs
1202 - Added RAPI resources:
1204 - ``/2/instances/[instance_name]/info``
1205 - ``/2/instances/[instance_name]/replace-disks``
1206 - ``/2/nodes/[node_name]/evacuate``
1207 - ``/2/nodes/[node_name]/migrate``
1208 - ``/2/nodes/[node_name]/role``
1209 - ``/2/nodes/[node_name]/storage``
1210 - ``/2/nodes/[node_name]/storage/modify``
1211 - ``/2/nodes/[node_name]/storage/repair``
1213 - Added OpCodes to evacuate or migrate all instances on a node
1214 - Added new command to list storage elements on nodes (``gnt-node
1215 list-storage``) and modify them (``gnt-node modify-storage``)
1216 - Added new ssconf files with master candidate IP address
1217 (``ssconf_master_candidates_ips``), node primary IP address
1218 (``ssconf_node_primary_ips``) and node secondary IP address
1219 (``ssconf_node_secondary_ips``)
1220 - Added ``ganeti-confd`` and a client library to query the Ganeti
1221 configuration via UDP
1222 - Added ability to run hooks after cluster initialization and before
1224 - Added automatic mode for disk replace (``gnt-instance replace-disks
1226 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1227 after catastrophic data-loss
1228 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1230 - Added ``gnt-instance move`` command to move instances
1231 - Added ``gnt-cluster watcher`` command to control watcher
1232 - Added ``gnt-node powercycle`` command to powercycle nodes
1233 - Added new job status field ``lock_status``
1234 - Added parseable error codes to cluster verification (``gnt-cluster
1235 verify --error-codes``) and made output less verbose (use
1236 ``--verbose`` to restore previous behaviour)
1237 - Added UUIDs to the main config entities (cluster, nodes, instances)
1238 - Added support for OS variants
1239 - Added support for hashed passwords in the Ganeti remote API users file
1241 - Added option to specify maximum timeout on instance shutdown
1242 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1243 - Added new helper script to start and stop Ganeti daemons
1244 (``daemon-util``), with the intent to reduce the work necessary to
1245 adjust Ganeti for non-Debian distributions and to start/stop daemons
1247 - Added more unittests
1248 - Fixed critical bug in ganeti-masterd startup
1249 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1250 now customisable at the cluster level for both the KVM and Xen
1251 hypervisors using the new ``migration_port`` parameter
1252 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1253 reinstalling an instance
1254 - Allowed ``@`` in tag names
1255 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1256 - Many documentation updates
1257 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1258 - ``gnt-instance reinstall`` can now reinstall multiple instances
1259 - Updated many command line parameters
1260 - Introduced new OS API version 15
1261 - No longer support a default hypervisor
1262 - Treat virtual LVs as inexistent
1263 - Improved job locking logic to reduce lock contention
1264 - Match instance and node names case insensitively
1265 - Reimplemented bash completion script to be more complete
1272 *(Released Thu, 4 Feb 2010)*
1274 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1275 - Fix a string formatting bug
1276 - Improve safety of the code in some error paths
1277 - Improve data validation in the master of values returned from nodes
1283 *(Released Thu, 17 Dec 2009)*
1285 - Fix security issue due to missing validation of iallocator names; this
1286 allows local and remote execution of arbitrary executables
1287 - Fix failure of gnt-node list during instance removal
1288 - Ship the RAPI documentation in the archive
1294 *(Released Wed, 30 Sep 2009)*
1296 - Fixed many wrong messages
1297 - Fixed a few bugs related to the locking library
1298 - Fixed MAC checking at instance creation time
1299 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1300 - Fixed a few issues related to signal handling in both daemons and
1302 - Fixed the example startup script provided
1303 - Fixed insserv dependencies in the example startup script (patch from
1305 - Fixed handling of drained nodes in the iallocator framework
1306 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1308 - Fixed error related to invalid job IDs in job polling
1309 - Fixed job/opcode persistence on unclean master shutdown
1310 - Fixed handling of partial job processing after unclean master
1312 - Fixed error reporting from LUs, previously all errors were converted
1313 into execution errors
1314 - Fixed error reporting from burnin
1315 - Decreased significantly the memory usage of the job queue
1316 - Optimised slightly multi-job submission
1317 - Optimised slightly opcode loading
1318 - Backported the multi-job submit framework from the development
1319 branch; multi-instance start and stop should be faster
1320 - Added script to clean archived jobs after 21 days; this will reduce
1321 the size of the queue directory
1322 - Added some extra checks in disk size tracking
1323 - Added an example ethers hook script
1324 - Added a cluster parameter that prevents Ganeti from modifying of
1326 - Added more node information to RAPI responses
1327 - Added a ``gnt-job watch`` command that allows following the ouput of a
1329 - Added a bind-address option to ganeti-rapi
1330 - Added more checks to the configuration verify
1331 - Enhanced the burnin script such that some operations can be retried
1333 - Converted instance reinstall to multi-instance model
1339 *(Released Fri, 7 Aug 2009)*
1341 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1342 to allow using the pre-2.0.2 behaviour in activation, if any existing
1343 instances have mismatched disk sizes in the configuration
1344 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1345 any configuration mismatches for disk sizes
1346 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1347 failover to work on two-node clusters
1348 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1350 - Fixed detection of OS script errors in ``gnt-backup export``
1351 - Fixed exit code of ``gnt-backup export``
1357 *(Released Fri, 17 Jul 2009)*
1359 - Added experimental support for stripped logical volumes; this should
1360 enhance performance but comes with a higher complexity in the block
1361 device handling; stripping is only enabled when passing
1362 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1363 affected even in the non-stripped mode
1364 - Improved resiliency against transient failures at the end of DRBD
1365 resyncs, and in general of DRBD resync checks
1366 - Fixed a couple of issues with exports and snapshot errors
1367 - Fixed a couple of issues in instance listing
1368 - Added display of the disk size in ``gnt-instance info``
1369 - Fixed checking for valid OSes in instance creation
1370 - Fixed handling of the "vcpus" parameter in instance listing and in
1371 general of invalid parameters
1372 - Fixed http server library, and thus RAPI, to handle invalid
1373 username/password combinations correctly; this means that now they
1374 report unauthorized for queries too, not only for modifications,
1375 allowing earlier detect of configuration problems
1376 - Added a new "role" node list field, equivalent to the master/master
1377 candidate/drained/offline flags combinations
1378 - Fixed cluster modify and changes of candidate pool size
1379 - Fixed cluster verify error messages for wrong files on regular nodes
1380 - Fixed a couple of issues with node demotion from master candidate role
1381 - Fixed node readd issues
1382 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1383 - Added a new ``--no-voting`` option for masterfailover to fix failover
1384 on two-nodes clusters when the former master node is unreachable
1385 - Added instance reinstall over RAPI
1391 *(Released Tue, 16 Jun 2009)*
1393 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1394 allow re-adding the start in single-user option (regression from 1.2)
1395 - the watcher writes the instance status to a file, to allow monitoring
1396 to report the instance status (from the master) based on cached
1397 results of the watcher's queries; while this can get stale if the
1398 watcher is being locked due to other work on the cluster, this is
1399 still an improvement
1400 - the watcher now also restarts the node daemon and the rapi daemon if
1402 - fixed the watcher to handle full and drained queue cases
1403 - hooks export more instance data in the environment, which helps if
1404 hook scripts need to take action based on the instance's properties
1405 (no longer need to query back into ganeti)
1406 - instance failovers when the instance is stopped do not check for free
1407 RAM, so that failing over a stopped instance is possible in low memory
1409 - rapi uses queries for tags instead of jobs (for less job traffic), and
1410 for cluster tags it won't talk to masterd at all but read them from
1412 - a couple of error handling fixes in RAPI
1413 - drbd handling: improved the error handling of inconsistent disks after
1414 resync to reduce the frequency of "there are some degraded disks for
1415 this instance" messages
1416 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1417 error handling path called a wrong function name)
1423 *(Released Wed, 27 May 2009)*
1425 - no changes from rc5
1428 Version 2.0 release candidate 5
1429 -------------------------------
1431 *(Released Wed, 20 May 2009)*
1433 - fix a couple of bugs (validation, argument checks)
1434 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1435 - some small improvements to RAPI and IAllocator
1436 - make watcher automatically start the master daemon if down
1439 Version 2.0 release candidate 4
1440 -------------------------------
1442 *(Released Mon, 27 Apr 2009)*
1444 - change the OS list to not require locks; this helps with big clusters
1445 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1446 volume group is broken
1447 - ``gnt-instance info``, without any arguments, doesn't run for all
1448 instances anymore; either pass ``--all`` or pass the desired
1449 instances; this helps against mistakes on big clusters where listing
1450 the information for all instances takes a long time
1451 - miscellaneous doc and man pages fixes
1454 Version 2.0 release candidate 3
1455 -------------------------------
1457 *(Released Wed, 8 Apr 2009)*
1459 - Change the internal locking model of some ``gnt-node`` commands, in
1460 order to reduce contention (and blocking of master daemon) when
1461 batching many creation/reinstall jobs
1462 - Fixes to Xen soft reboot
1463 - No longer build documentation at build time, instead distribute it in
1464 the archive, in order to reduce the need for the whole docbook/rst
1468 Version 2.0 release candidate 2
1469 -------------------------------
1471 *(Released Fri, 27 Mar 2009)*
1473 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1474 - Fix watcher startup sequence, improves the behaviour of busy clusters
1475 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1476 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1477 bind address and other places
1478 - Some documentation fixes and updates
1481 Version 2.0 release candidate 1
1482 -------------------------------
1484 *(Released Mon, 2 Mar 2009)*
1486 - More documentation updates, now all docs should be more-or-less
1488 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1490 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1492 - Fix an issue related to $libdir/run/ganeti and cluster creation
1498 *(Released Thu, 19 Feb 2009)*
1500 - Xen PVM and KVM have switched the default value for the instance root
1501 disk to the first partition on the first drive, instead of the whole
1502 drive; this means that the OS installation scripts must be changed
1504 - Man pages have been updated
1505 - RAPI has been switched by default to HTTPS, and the exported functions
1506 should all work correctly
1507 - RAPI v1 has been removed
1508 - Many improvements to the KVM hypervisor
1509 - Block device errors are now better reported
1510 - Many other bugfixes and small improvements
1516 *(Released Mon, 26 Jan 2009)*
1518 - Version 2 is a general rewrite of the code and therefore the
1519 differences are too many to list, see the design document for 2.0 in
1520 the ``doc/`` subdirectory for more details
1521 - In this beta version there is not yet a migration path from 1.2 (there
1522 will be one in the final 2.0 release)
1523 - A few significant changes are:
1525 - all commands are executed by a daemon (``ganeti-masterd``) and the
1526 various ``gnt-*`` commands are just front-ends to it
1527 - all the commands are entered into, and executed from a job queue,
1528 see the ``gnt-job(8)`` manpage
1529 - the RAPI daemon supports read-write operations, secured by basic
1530 HTTP authentication on top of HTTPS
1531 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1532 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1533 to migrate to DRBD 8 first while still running Ganeti 1.2)
1534 - DRBD devices are using statically allocated minor numbers, which
1535 will be assigned to existing instances during the migration process
1536 - there is support for both Xen PVM and Xen HVM instances running on
1538 - KVM virtualization is supported too
1539 - file-based storage has been implemented, which means that it is
1540 possible to run the cluster without LVM and DRBD storage, for
1541 example using a shared filesystem exported from shared storage (and
1542 still have live migration)
1548 *(Released Tue, 13 Jan 2009)*
1550 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1551 - Reuse the old instance mac address by default on instance import, if
1552 the instance name is the same.
1553 - Handle situations in which the node info rpc returns incomplete
1555 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1556 - Improved version of batcher:
1558 - state file support
1559 - instance mac address support
1560 - support for HVM clusters/instances
1562 - Add an option to show the number of cpu sockets and nodes in
1564 - Support OSes that handle more than one version of the OS api (but do
1565 not change the current API in any other way)
1566 - Fix ``gnt-node migrate``
1567 - ``gnt-debug`` man page
1568 - Fixes various more typos and small issues
1569 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1575 *(Released Wed, 24 Sep 2008)*
1577 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1578 type of disk exported to fully virtualized instances.
1579 - provide access to the serial console of HVM instances
1580 - instance auto_balance flag, set by default. If turned off it will
1581 avoid warnings on cluster verify if there is not enough memory to fail
1582 over an instance. in the future it will prevent automatically failing
1583 it over when we will support that.
1584 - batcher tool for instance creation, see ``tools/README.batcher``
1585 - ``gnt-instance reinstall --select-os`` to interactively select a new
1586 operating system when reinstalling an instance.
1587 - when changing the memory amount on instance modify a check has been
1588 added that the instance will be able to start. also warnings are
1589 emitted if the instance will not be able to fail over, if auto_balance
1591 - documentation fixes
1592 - sync fields between ``gnt-instance list/modify/add/import``
1593 - fix a race condition in drbd when the sync speed was set after giving
1594 the device a remote peer.
1600 *(Released Tue, 22 Jul 2008)*
1602 - note: the allowed size and number of tags per object were reduced
1603 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1604 - fixed twisted 8.x compatibility
1605 - fixed ``gnt-instance replace-disks`` with iallocator
1606 - add TCP keepalives on twisted connections to detect restarted nodes
1607 - disk increase support, see ``gnt-instance grow-disk``
1608 - implement bulk node/instance query for RAPI
1609 - add tags in node/instance listing (optional)
1610 - experimental migration (and live migration) support, read the man page
1611 for ``gnt-instance migrate``
1612 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1613 has some small improvements in handling its state file
1619 *(Released Fri, 13 Jun 2008)*
1621 - Experimental readonly, REST-based remote API implementation;
1622 automatically started on master node, TCP port 5080, if enabled by
1623 ``--enable-rapi`` parameter to configure script.
1624 - Instance allocator support. Add and import instance accept a
1625 ``--iallocator`` parameter, and call that instance allocator to decide
1626 which node to use for the instance. The iallocator document describes
1627 what's expected from an allocator script.
1628 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1629 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1630 node is lost there is still enough memory to fail over the instances
1632 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1633 ``gnt-cluster verify``, to check for site-specific compliance. All the
1634 hooks will run, and their output, if any, will be displayed. Any
1635 failing hook will make the verification return an error value.
1636 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1637 primary and secondary interfaces
1638 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1639 that is still declared as part of the cluster and has failed.
1640 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1641 specifying output fields, that just adds the chosen fields to the
1643 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1644 export from the filesystem.
1645 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1646 have been added. Using them you can enable/disable acpi and pae
1647 support, and specify a path for a cd image to be exported to the
1648 instance. These parameters as the name suggest only work on HVM
1650 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1651 PAE support will be set to the previously hardcoded values, but the
1652 (previously hardcoded) path to the CDROM ISO image will be unset and
1653 if required, needs to be set manually with ``gnt-instance modify``
1655 - The address to which an instance's VNC console is bound is now
1656 selectable per-instance, rather than being cluster wide. Of course
1657 this only applies to instances controlled via VNC, so currently just
1658 applies to HVM clusters.
1664 *(Released Mon, 18 Feb 2008)*
1666 - more tweaks to the disk activation code (especially helpful for DRBD)
1667 - change the default ``gnt-instance list`` output format, now there is
1668 one combined status field (see the manpage for the exact values this
1670 - some more fixes for the mac export to hooks change
1671 - make Ganeti not break with DRBD 8.2.x (which changed the version
1672 format in ``/proc/drbd``) (issue 24)
1673 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1674 template, allowing migration from DRBD0.7+MD to DRBD8
1680 *(Released Wed, 30 Jan 2008)*
1682 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1684 - add command aliases infrastructure and a few aliases
1685 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1686 man pages and the ``--help`` option of ``gnt-node
1687 list``/``gnt-instance list``
1688 - fix ``gnt-backup list`` with down nodes (issue 21)
1689 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1690 - fix the dist archive and add a check for including svn/git files in
1692 - some developer-related changes: improve the burnin and the QA suite,
1693 add an upload script for testing during development
1699 *(Released Wed, 16 Jan 2008)*
1701 - experimental HVM support, read the install document, section
1702 "Initializing the cluster"
1703 - allow for the PVM hypervisor per-instance kernel and initrd paths
1704 - add a new command ``gnt-cluster verify-disks`` which uses a new
1705 algorithm to improve the reconnection of the DRBD pairs if the device
1706 on the secondary node has gone away
1707 - make logical volume code auto-activate LVs at disk activation time
1708 - slightly improve the speed of activating disks
1709 - allow specification of the MAC address at instance creation time, and
1710 changing it later via ``gnt-instance modify``
1711 - fix handling of external commands that generate lots of output on
1713 - update documentation with regard to minimum version of DRBD8 supported
1719 *(Released Tue, 4 Dec 2007)*
1721 - Log the ``xm create`` output to the node daemon log on failure (to
1722 help diagnosing the error)
1723 - In debug mode, log all external commands output if failed to the logs
1724 - Change parsing of lvm commands to ignore stderr
1730 *(Released Wed, 28 Nov 2007)*
1732 - Another round of updates to the DRBD 8 code to deal with more failures
1733 in the replace secondary node operation
1734 - Some more logging of failures in disk operations (lvm, drbd)
1735 - A few documentation updates
1742 *(Released Tue, 13 Nov 2007)*
1744 - Change configuration file format from Python's Pickle to JSON.
1745 Upgrading is possible using the cfgupgrade utility.
1746 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1747 faster replace disks and is more stable (DRBD 8 has many improvements
1748 compared to DRBD 0.7)
1749 - Added command line tags support (see man pages for ``gnt-instance``,
1750 ``gnt-node``, ``gnt-cluster``)
1751 - Added instance rename support
1752 - Added multi-instance startup/shutdown
1753 - Added cluster rename support
1754 - Added ``gnt-node evacuate`` to simplify some node operations
1755 - Added instance reboot operation that can speedup reboot as compared to
1757 - Soften the requirement that hostnames are in FQDN format
1758 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1760 - Removed dependency on debian's patched fping that uses the
1761 non-standard ``-S`` option
1762 - Now the OS definitions are searched for in multiple, configurable
1763 paths (easier for distros to package)
1764 - Some changes to the hooks infrastructure (especially the new
1765 post-configuration update hook)
1766 - Other small bugfixes
1768 .. vim: set textwidth=72 syntax=rst :