7 *(Released Wed, 31 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.
88 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
90 Node group improvements
91 ~~~~~~~~~~~~~~~~~~~~~~~
93 - ``gnt-cluster verify`` has been modified to check groups separately,
94 thereby improving performance.
95 - Node group support has been added to ``gnt-cluster verify-disks``,
96 which now operates per node group.
97 - Watcher has been changed to work better with node groups.
99 - One process and state file per node group.
100 - Slow watcher in one group doesn't block other group's watcher.
102 - Added new command, ``gnt-group evacuate``, to move all instances in a
103 node group to other groups.
104 - Added ``gnt-instance change-group`` to move an instance to another
106 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
107 per-group operations.
108 - Node groups can be tagged.
109 - Some operations switch from an exclusive to a shared lock as soon as
111 - Instance's primary and secondary nodes' groups are now available as
112 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
113 and ``snodes.group.uuid``).
118 - Numerous updates to documentation and manpages.
120 - :doc:`RAPI <rapi>` documentation now has detailed parameter
123 - A lockset's internal lock is now also visible in lock monitor.
124 - Log messages from job queue workers now contain information about the
125 opcode they're processing.
126 - ``gnt-instance console`` no longer requires the instance lock.
127 - A short delay when waiting for job changes reduces the number of LUXI
128 requests significantly.
129 - DRBD metadata volumes are overwritten with zeros during disk creation.
135 *(Released Mon, 22 Aug 2011)*
137 This was the second beta release of the 2.5 series.
143 *(Released Fri, 12 Aug 2011)*
145 This was the first beta release of the 2.5 series.
151 *(Released Tue, 23 Aug 2011)*
155 - Fixed documentation for importing with ``--src-dir`` option
156 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
157 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
163 *(Released Fri, 5 Aug 2011)*
165 Many bug-fixes and a few small features:
167 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
168 issues when you tried to add an instance with two MAC addresses in one
170 - KVM: fixed per-instance stored UID value
171 - KVM: configure bridged NICs at migration start
172 - KVM: Fix a bug where instance will not start with never KVM versions
174 - Added OS search path to ``gnt-cluster info``
175 - Fixed an issue with ``file_storage_dir`` where you were forced to
176 provide an absolute path, but the documentation states it is a
177 relative path, the documentation was right
178 - Added a new parameter to instance stop/start called ``--no-remember``
179 that will make the state change to not be remembered
180 - Implemented ``no_remember`` at RAPI level
181 - Improved the documentation
182 - Node evacuation: don't call IAllocator if node is already empty
183 - Fixed bug in DRBD8 replace disks on current nodes
184 - Fixed bug in recreate-disks for DRBD instances
185 - Moved assertion checking locks in ``gnt-instance replace-disks``
186 causing it to abort with not owning the right locks for some situation
187 - Job queue: Fixed potential race condition when cancelling queued jobs
188 - Fixed off-by-one bug in job serial generation
189 - ``gnt-node volumes``: Fix instance names
190 - Fixed aliases in bash completion
191 - Fixed a bug in reopening log files after being sent a SIGHUP
192 - Added a flag to burnin to allow specifying VCPU count
193 - Bugfixes to non-root Ganeti configuration
199 *(Released Thu, 12 May 2011)*
201 Many bug-fixes and a few new small features:
203 - Fixed a bug related to log opening failures
204 - Fixed a bug in instance listing with orphan instances
205 - Fixed a bug which prevented resetting the cluster-level node parameter
206 ``oob_program`` to the default
207 - Many fixes related to the ``cluster-merge`` tool
208 - Fixed a race condition in the lock monitor, which caused failures
209 during (at least) creation of many instances in parallel
210 - Improved output for gnt-job info
211 - Removed the quiet flag on some ssh calls which prevented debugging
213 - Improved the N+1 failure messages in cluster verify by actually
214 showing the memory values (needed and available)
215 - Increased lock attempt timeouts so that when executing long operations
216 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
217 too early and thus prevent the use of the 'fair' mechanism
218 - Changed instance query data (``gnt-instance info``) to not acquire
219 locks unless needed, thus allowing its use on locked instance if only
220 static information is asked for
221 - Improved behaviour with filesystems that do not support rename on an
223 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
224 kept locks on all nodes during the wipe, which is unneeded
225 - Fixed ``gnt-watcher`` handling of errors during hooks execution
226 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
227 10GiB) which caused the wipe to fail right at the end in some cases
228 - Fixed master IP activation when doing master failover with no-voting
229 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
230 the master node itself
231 - Fixed potential data-loss in under disk full conditions, where Ganeti
232 wouldn't check correctly the return code and would consider
233 partially-written files 'correct'
234 - Fixed bug related to multiple VGs and DRBD disk replacing
235 - Added new disk parameter ``metavg`` that allows placement of the meta
236 device for DRBD in a different volume group
237 - Fixed error handling in the node daemon when the system libc doesn't
238 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
239 - Fixed lock release during replace-disks, which kept cluster-wide locks
240 when doing disk replaces with an iallocator script
241 - Added check for missing bridges in cluster verify
242 - Handle EPIPE errors while writing to the terminal better, so that
243 piping the output to e.g. ``less`` doesn't cause a backtrace
244 - Fixed rare case where a ^C during Luxi calls could have been
245 interpreted as server errors, instead of simply terminating
246 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
248 - Added a few more parameters to the KVM hypervisor, allowing a second
249 CDROM, custom disk type for CDROMs and a floppy image
250 - Removed redundant message in instance rename when the name is given
252 - Added option to ``gnt-instance recreate-disks`` to allow creating the
253 disks on new nodes, allowing recreation when the original instance
254 nodes are completely gone
255 - Added option when converting disk templates to DRBD to skip waiting
256 for the resync, in order to make the instance available sooner
257 - Added two new variables to the OS scripts environment (containing the
259 - Made the root_path and optional parameter for the xen-pvm hypervisor,
260 to allow use of ``pvgrub`` as bootloader
261 - Changed the instance memory modifications to only check out-of-memory
262 conditions on memory increases, and turned the secondary node warnings
263 into errors (they can still be overridden via ``--force``)
264 - Fixed the handling of a corner case when the Python installation gets
265 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
266 to execute a command that doesn't exist
267 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
268 node of the instance returned failures during instance shutdown; this
269 adds the option ``--ignore-consistency`` to gnt-instance move
271 And as usual, various improvements to the error messages, documentation
277 *(Released Wed, 09 Mar 2011)*
279 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
280 the RAPI users file if run twice (even with ``--dry-run``).
282 The release fixes that bug (nothing else changed).
288 *(Released Mon, 07 Mar 2011)*
290 Final 2.4.0 release. Just a few small fixes:
292 - Fixed RAPI node evacuate
293 - Fixed the kvm-ifup script
294 - Fixed internal error handling for special job cases
295 - Updated man page to specify the escaping feature for options
301 *(Released Mon, 28 Feb 2011)*
303 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
304 that this feature wiped the disks of the wrong instance, leading to loss
309 - Fixed title of query field containing instance name
310 - Expanded the glossary in the documentation
311 - Fixed one unittest (internal issue)
317 *(Released Mon, 21 Feb 2011)*
319 A number of bug fixes plus just a couple functionality changes.
321 On the user-visible side, the ``gnt-* list`` command output has changed
322 with respect to "special" field states. The current rc1 style of display
323 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
324 the default output mode special fields states are displayed as follows:
326 - Offline resource: ``*``
327 - Unavailable/not applicable: ``-``
328 - Data missing (RPC failure): ``?``
329 - Unknown field: ``??``
331 Another user-visible change is the addition of ``--force-join`` to
336 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
337 - Fixed regression in RAPI/instance reinstall where all parameters were
338 required (instead of optional)
339 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
340 - Fixed iallocator usage (offline nodes were not considered offline)
341 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
342 - Fixed hypervisor and OS parameter validation with respect to
344 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
346 - Fixed ``tools/listrunner`` with respect to agent-based usage
352 *(Released Fri, 4 Feb 2011)*
354 Many changes and fixes since the beta1 release. While there were some
355 internal changes, the code has been mostly stabilised for the RC
358 Note: the dumb allocator was removed in this release, as it was not kept
359 up-to-date with the IAllocator protocol changes. It is recommended to
360 use the ``hail`` command from the ganeti-htools package.
362 Note: the 2.4 and up versions of Ganeti are not compatible with the
363 0.2.x branch of ganeti-htools. You need to upgrade to
364 ganeti-htools-0.3.0 (or later).
366 Regressions fixed from 2.3
367 ~~~~~~~~~~~~~~~~~~~~~~~~~~
369 - Fixed the ``gnt-cluster verify-disks`` command
370 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
372 - Fixed disk adoption breakage
373 - Fixed wrong headers in instance listing for field aliases
378 - Fixed corner case in KVM handling of NICs
379 - Fixed many cases of wrong handling of non-vm_capable nodes
380 - Fixed a bug where a missing instance symlink was not possible to
381 recreate with any ``gnt-*`` command (now ``gnt-instance
382 activate-disks`` does it)
383 - Fixed the volume group name as reported by ``gnt-cluster
385 - Increased timeouts for the import-export code, hopefully leading to
386 fewer aborts due network or instance timeouts
387 - Fixed bug in ``gnt-node list-storage``
388 - Fixed bug where not all daemons were started on cluster
389 initialisation, but only at the first watcher run
390 - Fixed many bugs in the OOB implementation
391 - Fixed watcher behaviour in presence of instances with offline
393 - Fixed instance list output for instances running on the wrong node
394 - a few fixes to the cluster-merge tool, but it still cannot merge
395 multi-node groups (currently it is not recommended to use this tool)
401 - Improved network configuration for the KVM hypervisor
402 - Added e1000 as a supported NIC for Xen-HVM
403 - Improved the lvmstrap tool to also be able to use partitions, as
404 opposed to full disks
405 - Improved speed of disk wiping (the cluster parameter
406 ``prealloc_wipe_disks``, so that it has a low impact on the total time
407 of instance creations
408 - Added documentation for the OS parameters
409 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
410 hypervisor is not responding
411 - Added display of blacklisted and hidden OS information in
413 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
414 NIC and node parameters, which might create problems with currently
415 invalid (but undetected) configuration files, but prevents validation
416 failures when unrelated parameters are modified
417 - Changed cluster initialisation to wait for the master daemon to become
419 - Expanded the RAPI interface:
421 - Added config redistribution resource
422 - Added activation/deactivation of instance disks
423 - Added export of console information
425 - Implemented log file reopening on SIGHUP, which allows using
426 logrotate(8) for the Ganeti log files
427 - Added a basic OOB helper script as an example
433 *(Released Fri, 14 Jan 2011)*
438 - Fixed timezone issues when formatting timestamps
439 - Added support for node groups, available via ``gnt-group`` and other
441 - Added out-of-band framework and management, see :doc:`design
442 document <design-oob>`
443 - Removed support for roman numbers from ``gnt-node list`` and
444 ``gnt-instance list``.
445 - Allowed modification of master network interface via ``gnt-cluster
446 modify --master-netdev``
447 - Accept offline secondaries while shutting down instance disks
448 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
449 - Added support for multiple LVM volume groups
450 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
452 - Added commands to list available fields:
454 - ``gnt-node list-fields``
455 - ``gnt-group list-fields``
456 - ``gnt-instance list-fields``
458 - Updated documentation and man pages
463 - Moved ``rapi_users`` file into separate directory, now named
464 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
466 - Added new tool for running commands on many machines,
467 ``tools/ganeti-listrunner``
468 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
469 improving the ``gnt-instance console`` output
470 - Allowed customisation of disk index separator at ``configure`` time
471 - Export node group allocation policy to :doc:`iallocator <iallocator>`
472 - Added support for non-partitioned md disks in ``lvmstrap``
473 - Added script to gracefully power off KVM instances
474 - Split ``utils`` module into smaller parts
475 - Changed query operations to return more detailed information, e.g.
476 whether an information is unavailable due to an offline node. To use
477 this new functionality, the LUXI call ``Query`` must be used. Field
478 information is now stored by the master daemon and can be retrieved
479 using ``QueryFields``. Instances, nodes and groups can also be queried
480 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
481 exposed via RAPI). The following commands make use of this
482 infrastructure change:
485 - ``gnt-group list-fields``
487 - ``gnt-node list-fields``
488 - ``gnt-instance list``
489 - ``gnt-instance list-fields``
490 - ``gnt-debug locks``
495 - New RAPI resources (see :doc:`rapi`):
499 - ``/2/groups/[group_name]``
500 - ``/2/groups/[group_name]/assign-nodes``
501 - ``/2/groups/[group_name]/modify``
502 - ``/2/groups/[group_name]/rename``
503 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
507 - Implemented ``no_install`` for instance creation
508 - Implemented OS parameters for instance reinstallation, allowing
509 use of special settings on reinstallation (e.g. for preserving data)
514 - Added IPv6 support in import/export
515 - Pause DRBD synchronization while wiping disks on instance creation
516 - Updated unittests and QA scripts
517 - Improved network parameters passed to KVM
518 - Converted man pages from docbook to reStructuredText
524 *(Released Mon, 20 Dec 2010)*
526 Released version 2.3.1~rc1 without any changes.
532 *(Released Wed, 1 Dec 2010)*
534 - impexpd: Disable OpenSSL compression in socat if possible (backport
535 from master, commit e90739d625b, see :doc:`installation guide
536 <install-quick>` for details)
537 - Changed unittest coverage report to exclude test scripts
538 - Added script to check version format
544 *(Released Wed, 1 Dec 2010)*
546 Released version 2.3.0~rc1 without any changes.
552 *(Released Fri, 19 Nov 2010)*
554 A number of bugfixes and documentation updates:
556 - Update ganeti-os-interface documentation
557 - Fixed a bug related to duplicate MACs or similar items which should be
559 - Fix breakage in OS state modify
560 - Reinstall instance: disallow offline secondaries (fixes bug related to
561 OS changing but reinstall failing)
562 - plus all the other fixes between 2.2.1 and 2.2.2
568 *(Released Tue, 2 Nov 2010)*
570 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
571 - Fixed master failover race with watcher
572 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
574 - Accept previously stopped instance for export with instance removal
575 - Simplify and extend the environment variables for instance OS scripts
576 - Added new node flags, ``master_capable`` and ``vm_capable``
577 - Added optional instance disk wiping prior during allocation. This is a
578 cluster-wide option and can be set/modified using
579 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
580 - Added IPv6 support, see :doc:`design document <design-2.3>` and
582 - Added a new watcher option (``--ignore-pause``)
583 - Added option to ignore offline node on instance start/stop
584 (``--ignore-offline``)
585 - Allow overriding OS parameters with ``gnt-instance reinstall``
586 - Added ability to change node's secondary IP address using ``gnt-node
588 - Implemented privilege separation for all daemons except
589 ``ganeti-noded``, see ``configure`` options
590 - Complain if an instance's disk is marked faulty in ``gnt-cluster
592 - Implemented job priorities (see ``ganeti(7)`` manpage)
593 - Ignore failures while shutting down instances during failover from
595 - Exit daemon's bootstrap process only once daemon is ready
596 - Export more information via ``LUInstanceQuery``/remote API
597 - Improved documentation, QA and unittests
598 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
599 restart if the file was created or changed
600 - Added LUXI protocol version sent with each request and response,
601 allowing detection of server/client mismatches
602 - Moved the Python scripts among gnt-* and ganeti-* into modules
603 - Moved all code related to setting up SSH to an external script,
605 - Infrastructure changes for node group support in future versions
611 *(Released Fri, 19 Nov 2010)*
613 A few small bugs fixed, and some improvements to the build system:
615 - Fix documentation regarding conversion to drbd
616 - Fix validation of parameters in cluster modify (``gnt-cluster modify
618 - Fix error handling in node modify with multiple changes
619 - Allow remote imports without checked names
625 *(Released Tue, 19 Oct 2010)*
627 - Disable SSL session ID cache in RPC client
633 *(Released Thu, 14 Oct 2010)*
635 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
636 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
638 - Fix problems with interaction between Curl and Python's HTTP server,
639 resulting in increased speed in many RPC calls
640 - Improve our release script to prevent breakage with older aclocal and
647 *(Released Thu, 7 Oct 2010)*
649 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
650 value retrieved from master
651 - Added support for blacklisted or hidden OS definitions
652 - Added simple lock monitor (accessible via (``gnt-debug locks``)
653 - Added support for -mem-path in KVM hypervisor abstraction layer
654 - Allow overriding instance parameters in tool for inter-cluster
655 instance moves (``tools/move-instance``)
656 - Improved opcode summaries (e.g. in ``gnt-job list``)
657 - Improve consistency of OS listing by sorting it
658 - Documentation updates
664 *(Released Fri, 8 Oct 2010)*
666 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
672 *(Released Mon, 4 Oct 2010)*
674 - Fixed regression in ``gnt-instance rename``
680 *(Released Wed, 22 Sep 2010)*
682 - Fixed OS_VARIANT variable for OS scripts
683 - Fixed cluster tag operations via RAPI
684 - Made ``setup-ssh`` exit with non-zero code if an error occurred
685 - Disabled RAPI CA checks in watcher
691 *(Released Mon, 23 Aug 2010)*
693 - Support DRBD versions of the format "a.b.c.d"
695 - Re-introduce support for usage from multiple threads in RAPI client
696 - Instance renames and modify via RAPI
697 - Work around race condition between processing and archival in job
699 - Mark opcodes following failed one as failed, too
700 - Job field ``lock_status`` was removed due to difficulties making it
701 work with the changed job queue in Ganeti 2.2; a better way to monitor
702 locks is expected for a later 2.2.x release
703 - Fixed dry-run behaviour with many commands
704 - Support ``ssh-agent`` again when adding nodes
705 - Many additional bugfixes
711 *(Released Fri, 30 Jul 2010)*
713 Important change: the internal RPC mechanism between Ganeti nodes has
714 changed from using a home-grown http library (based on the Python base
715 libraries) to use the PycURL library. This requires that PycURL is
716 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
717 against GnuTLS by default. cURL's support for GnuTLS had known issues
718 before cURL 7.21.0 and we recommend using the latest cURL release or
719 linking against OpenSSL. Most other distributions already link PycURL
720 and cURL against OpenSSL. The command::
722 python -c 'import pycurl; print pycurl.version'
724 can be used to determine the libraries PycURL and cURL are linked
727 Other significant changes:
729 - Rewrote much of the internals of the job queue, in order to achieve
730 better parallelism; this decouples job query operations from the job
731 processing, and it should allow much nicer behaviour of the master
732 daemon under load, and it also has uncovered some long-standing bugs
733 related to the job serialisation (now fixed)
734 - Added a default iallocator setting to the cluster parameters,
735 eliminating the need to always pass nodes or an iallocator for
736 operations that require selection of new node(s)
737 - Added experimental support for the LXC virtualization method
738 - Added support for OS parameters, which allows the installation of
739 instances to pass parameter to OS scripts in order to customise the
741 - Added a hypervisor parameter controlling the migration type (live or
742 non-live), since hypervisors have various levels of reliability; this
743 has renamed the 'live' parameter to 'mode'
744 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
745 logical volumes, meaning that cluster verify will ignore them and not
746 flag their presence as errors
747 - The watcher will now reset the error count for failed instances after
748 8 hours, thus allowing self-healing if the problem that caused the
749 instances to be down/fail to start has cleared in the meantime
750 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
751 check for, and warn, if the drbd module parameter ``usermode_helper``
752 is not consistent with the cluster-wide setting; this is needed to
753 make diagnose easier of failed drbd creations
754 - Started adding base IPv6 support, but this is not yet
755 enabled/available for use
756 - Rename operations (cluster, instance) will now return the new name,
757 which is especially useful if a short name was passed in
758 - Added support for instance migration in RAPI
759 - Added a tool to pre-configure nodes for the SSH setup, before joining
760 them to the cluster; this will allow in the future a simplified model
761 for node joining (but not yet fully enabled in 2.2); this needs the
762 paramiko python library
763 - Fixed handling of name-resolving errors
764 - Fixed consistency of job results on the error path
765 - Fixed master-failover race condition when executed multiple times in
767 - Fixed many bugs related to the job queue (mostly introduced during the
768 2.2 development cycle, so not all are impacting 2.1)
769 - Fixed instance migration with missing disk symlinks
770 - Fixed handling of unknown jobs in ``gnt-job archive``
771 - And many other small fixes/improvements
775 - Enhanced both the unittest and the QA coverage
776 - Switched the opcode validation to a generic model, and extended the
777 validation to all opcode parameters
778 - Changed more parts of the code that write shell scripts to use the
780 - Switched the master daemon to use the asyncore library for the Luxi
787 *(Released Thu, 17 Jun 2010)*
789 - Added tool (``move-instance``) and infrastructure to move instances
790 between separate clusters (see :doc:`separate documentation
791 <move-instance>` and :doc:`design document <design-2.2>`)
792 - Added per-request RPC timeout
793 - RAPI now requires a Content-Type header for requests with a body (e.g.
794 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
795 :rfc:`2616` (HTTP/1.1), section 7.2.1)
796 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
798 - Implemented initial support for running Ganeti daemons as separate
799 users, see configure-time flags ``--with-user-prefix`` and
800 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
802 - Instances can be removed after export (``gnt-backup export
804 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
805 key (instead of 1024 bit)
806 - Added new cluster configuration file for cluster domain secret
807 - Import/export now use SSL instead of SSH
808 - Added support for showing estimated time when exporting an instance,
809 see the ``ganeti-os-interface(7)`` manpage and look for
816 *(Released Tue, 16 Nov 2010)*
818 Some more bugfixes. Unless critical bugs occur, this will be the last
821 - Fix case of MAC special-values
822 - Fix mac checker regex
823 - backend: Fix typo causing "out of range" error
824 - Add missing --units in gnt-instance list man page
830 *(Released Tue, 24 Aug 2010)*
833 - Don't ignore secondary node silently on non-mirrored disk templates
835 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
836 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
837 - Properly document the usb_mouse parameter
838 - Fix path in ganeti-rapi(8) (issue 116)
839 - Adjust error message when the ganeti user's .ssh directory is
841 - Add same-node-check when changing the disk template to drbd
847 *(Released Fri, 16 Jul 2010)*
850 - Add an option to only select some reboot types during qa/burnin.
851 (on some hypervisors consequent reboots are not supported)
852 - Fix infrequent race condition in master failover. Sometimes the old
853 master ip address would be still detected as up for a short time
854 after it was removed, causing failover to fail.
855 - Decrease mlockall warnings when the ctypes module is missing. On
856 Python 2.4 we support running even if no ctypes module is installed,
857 but we were too verbose about this issue.
858 - Fix building on old distributions, on which man doesn't have a
860 - Fix RAPI not to ignore the MAC address on instance creation
861 - Implement the old instance creation format in the RAPI client.
867 *(Released Thu, 01 Jul 2010)*
869 A small bugfix release:
870 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
871 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
872 a non-existing option
873 - Fix instance create: the --force-variant option was ignored
874 - Improve pylint 0.21 compatibility and warnings with Python 2.6
875 - Fix modify node storage with non-FQDN arguments
876 - Fix RAPI client to authenticate under Python 2.6 when used
877 for more than 5 requests needing authentication
878 - Fix gnt-instance modify -t (storage) giving a wrong error message
879 when converting a non-shutdown drbd instance to plain
885 *(Released Fri, 18 Jun 2010)*
887 A small bugfix release:
889 - Fix live migration of KVM instances started with older Ganeti
890 versions which had fewer hypervisor parameters
891 - Fix gnt-instance grow-disk on down instances
892 - Fix an error-reporting bug during instance migration
893 - Better checking of the ``--net`` and ``--disk`` values, to avoid
894 silently ignoring broken ones
895 - Fix an RPC error reporting bug affecting, for example, RAPI client
897 - Fix bug triggered by different API version os-es on different nodes
898 - Fix a bug in instance startup with custom hvparams: OS level
899 parameters would fail to be applied.
900 - Fix the RAPI client under Python 2.6 (but more work is needed to
901 make it work completely well with OpenSSL)
902 - Fix handling of errors when resolving names from DNS
908 *(Released Thu, 3 Jun 2010)*
910 A medium sized development cycle. Some new features, and some
911 fixes/small improvements/cleanups.
916 The node deamon now tries to mlock itself into memory, unless the
917 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
918 its logs, and falls back to console logging. This allows emergency
919 features such as ``gnt-node powercycle`` to work even in the event of a
920 broken node disk (tested offlining the disk hosting the node's
921 filesystem and dropping its memory caches; don't try this at home)
923 KVM: add vhost-net acceleration support. It can be tested with a new
924 enough version of the kernel and of qemu-kvm.
926 KVM: Add instance chrooting feature. If you use privilege dropping for
927 your VMs you can also now force them to chroot to an empty directory,
928 before starting the emulated guest.
930 KVM: Add maximum migration bandwith and maximum downtime tweaking
931 support (requires a new-enough version of qemu-kvm).
933 Cluster verify will now warn if the master node doesn't have the master
936 Add a new (incompatible) instance creation request format to RAPI which
937 supports all parameters (previously only a subset was supported, and it
938 wasn't possible to extend the old format to accomodate all the new
939 features. The old format is still supported, and a client can check for
940 this feature, before using it, by checking for its presence in the
941 ``features`` RAPI resource.
943 Now with ancient latin support. Try it passing the ``--roman`` option to
944 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
945 (requires the python-roman module to be installed, in order to work).
950 As usual many internal code refactorings, documentation updates, and
953 - Lots of improvements and cleanups to the experimental Remote API
954 (RAPI) client library.
955 - A new unit test suite for the core daemon libraries.
956 - A fix to creating missing directories makes sure the umask is not
957 applied anymore. This enforces the same directory permissions
959 - Better handling terminating daemons with ctrl+c (used when running
960 them in debugging mode).
961 - Fix a race condition in live migrating a KVM instance, when stat()
962 on the old proc status file returned EINVAL, which is an unexpected
964 - Fixed manpage checking with newer man and utf-8 charachters. But now
965 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
971 *(Released Fri, 7 May 2010)*
973 Fix a bug which prevented untagged KVM instances from starting.
979 *(Released Fri, 7 May 2010)*
981 Another release with a long development cycle, during which many
982 different features were added.
987 The KVM hypervisor now can run the individual instances as non-root, to
988 reduce the impact of a VM being hijacked due to bugs in the
989 hypervisor. It is possible to run all instances as a single (non-root)
990 user, to manually specify a user for each instance, or to dynamically
991 allocate a user out of a cluster-wide pool to each instance, with the
992 guarantee that no two instances will run under the same user ID on any
995 An experimental RAPI client library, that can be used standalone
996 (without the other Ganeti libraries), is provided in the source tree as
997 ``lib/rapi/client.py``. Note this client might change its interface in
998 the future, as we iterate on its capabilities.
1000 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1001 replace the cluster's certificates and crypto keys. This might help in
1002 case they have been compromised, or have simply expired.
1004 A new disk option for instance creation has been added that allows one
1005 to "adopt" currently existing logical volumes, with data
1006 preservation. This should allow easier migration to Ganeti from
1007 unmanaged (or managed via other software) instances.
1009 Another disk improvement is the possibility to convert between redundant
1010 (DRBD) and plain (LVM) disk configuration for an instance. This should
1011 allow better scalability (starting with one node and growing the
1012 cluster, or shrinking a two-node cluster to one node).
1014 A new feature that could help with automated node failovers has been
1015 implemented: if a node sees itself as offline (by querying the master
1016 candidates), it will try to shutdown (hard) all instances and any active
1017 DRBD devices. This reduces the risk of duplicate instances if an
1018 external script automatically failovers the instances on such nodes. To
1019 enable this, the cluster parameter ``maintain_node_health`` should be
1020 enabled; in the future this option (per the name) will enable other
1021 automatic maintenance features.
1023 Instance export/import now will reuse the original instance
1024 specifications for all parameters; that means exporting an instance,
1025 deleting it and the importing it back should give an almost identical
1026 instance. Note that the default import behaviour has changed from
1027 before, where it created only one NIC; now it recreates the original
1030 Cluster verify has added a few new checks: SSL certificates validity,
1031 /etc/hosts consistency across the cluster, etc.
1036 As usual, many internal changes were done, documentation fixes,
1039 - Fixed cluster initialization with disabled cluster storage (regression
1040 introduced in 2.1.1)
1041 - File-based storage supports growing the disks
1042 - Fixed behaviour of node role changes
1043 - Fixed cluster verify for some corner cases, plus a general rewrite of
1044 cluster verify to allow future extension with more checks
1045 - Fixed log spamming by watcher and node daemon (regression introduced
1047 - Fixed possible validation issues when changing the list of enabled
1049 - Fixed cleanup of /etc/hosts during node removal
1050 - Fixed RAPI response for invalid methods
1051 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1052 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1053 from ide disks, etc.)
1054 - Allow OS changes without re-installation (to record a changed OS
1055 outside of Ganeti, or to allow OS renames)
1056 - Allow instance creation without OS installation (useful for example if
1057 the OS will be installed manually, or restored from a backup not in
1059 - Implemented option to make cluster ``copyfile`` use the replication
1061 - Added list of enabled hypervisors to ssconf (possibly useful for
1063 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1065 - A partial form of node re-IP is possible via node readd, which now
1066 allows changed node primary IP
1067 - Command line utilities now show an informational message if the job is
1069 - The logs of the master daemon now show the PID/UID/GID of the
1076 *(Released Fri, 12 Mar 2010)*
1078 During the 2.1.0 long release candidate cycle, a lot of improvements and
1079 changes have accumulated with were released later as 2.1.1.
1084 The node evacuate command (``gnt-node evacuate``) was significantly
1085 rewritten, and as such the IAllocator protocol was changed - a new
1086 request type has been added. This unfortunate change during a stable
1087 series is designed to improve performance of node evacuations; on
1088 clusters with more than about five nodes and which are well-balanced,
1089 evacuation should proceed in parallel for all instances of the node
1090 being evacuated. As such, any existing IAllocator scripts need to be
1091 updated, otherwise the above command will fail due to the unknown
1092 request. The provided "dumb" allocator has not been updated; but the
1093 ganeti-htools package supports the new protocol since version 0.2.4.
1095 Another important change is increased validation of node and instance
1096 names. This might create problems in special cases, if invalid host
1097 names are being used.
1099 Also, a new layer of hypervisor parameters has been added, that sits at
1100 OS level between the cluster defaults and the instance ones. This allows
1101 customisation of virtualization parameters depending on the installed
1102 OS. For example instances with OS 'X' may have a different KVM kernel
1103 (or any other parameter) than the cluster defaults. This is intended to
1104 help managing a multiple OSes on the same cluster, without manual
1105 modification of each instance's parameters.
1107 A tool for merging clusters, ``cluster-merge``, has been added in the
1108 tools sub-directory.
1113 - Improved the int/float conversions that should make the code more
1114 robust in face of errors from the node daemons
1115 - Fixed the remove node code in case of internal configuration errors
1116 - Fixed the node daemon behaviour in face of inconsistent queue
1117 directory (e.g. read-only file-system where we can't open the files
1119 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1120 now it either aborts cleanly or, if given the new "auto_promote"
1121 parameter, will automatically promote other nodes as needed
1122 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1123 completely prevent Ganeti from working
1124 - Fixed bug for instance export when not all disks were successfully
1126 - Fixed behaviour of node add when the new node is slow in starting up
1128 - Fixed handling of signals in the LUXI client, which should improve
1129 behaviour of command-line scripts
1130 - Added checks for invalid node/instance names in the configuration (now
1131 flagged during cluster verify)
1132 - Fixed watcher behaviour for disk activation errors
1133 - Fixed two potentially endless loops in http library, which led to the
1134 RAPI daemon hanging and consuming 100% CPU in some cases
1135 - Fixed bug in RAPI daemon related to hashed passwords
1136 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1138 - Enhanced compatibility with non-Debian OSes, but not using absolute
1139 path in some commands and allowing customisation of the ssh
1140 configuration directory
1141 - Fixed possible future issue with new Python versions by abiding to the
1142 proper use of ``__slots__`` attribute on classes
1143 - Added checks that should prevent directory traversal attacks
1144 - Many documentation fixes based on feedback from users
1149 - Added an "early_release" more for instance replace disks and node
1150 evacuate, where we release locks earlier and thus allow higher
1151 parallelism within the cluster
1152 - Added watcher hooks, intended to allow the watcher to restart other
1153 daemons (e.g. from the ganeti-nbma project), but they can be used of
1154 course for any other purpose
1155 - Added a compile-time disable for DRBD barriers, to increase
1156 performance if the administrator trusts the power supply or the
1157 storage system to not lose writes
1158 - Added the option of using syslog for logging instead of, or in
1159 addition to, Ganeti's own log files
1160 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1161 can indeed boot from a paravirtual NIC
1162 - Added a generic debug level for many operations; while this is not
1163 used widely yet, it allows one to pass the debug value all the way to
1165 - Enhanced the hooks environment for instance moves (failovers,
1166 migrations) where the primary/secondary nodes changed during the
1167 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1168 - Enhanced data validations for many user-supplied values; one important
1169 item is the restrictions imposed on instance and node names, which
1170 might reject some (invalid) host names
1171 - Add a configure-time option to disable file-based storage, if it's not
1172 needed; this allows greater security separation between the master
1173 node and the other nodes from the point of view of the inter-node RPC
1175 - Added user notification in interactive tools if job is waiting in the
1176 job queue or trying to acquire locks
1177 - Added log messages when a job is waiting for locks
1178 - Added filtering by node tags in instance operations which admit
1179 multiple instances (start, stop, reboot, reinstall)
1180 - Added a new tool for cluster mergers, ``cluster-merge``
1181 - Parameters from command line which are of the form ``a=b,c=d`` can now
1182 use backslash escapes to pass in values which contain commas,
1183 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1185 - For KVM, the instance name is the first parameter passed to KVM, so
1186 that it's more visible in the process list
1192 *(Released Tue, 2 Mar 2010)*
1194 Ganeti 2.1 brings many improvements with it. Major changes:
1196 - Added infrastructure to ease automated disk repairs
1197 - Added new daemon to export configuration data in a cheaper way than
1198 using the remote API
1199 - Instance NICs can now be routed instead of being associated with a
1201 - Improved job locking logic to reduce impact of jobs acquiring multiple
1202 locks waiting for other long-running jobs
1204 In-depth implementation details can be found in the Ganeti 2.1 design
1210 - Added chroot hypervisor
1211 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1213 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1214 ``bootloader_path`` and ``bootloader_args``)
1215 - Added the ``use_localtime`` option for the xen-hvm and kvm
1216 hypervisors, and the default value for this has changed to false (in
1217 2.0 xen-hvm always enabled it)
1218 - Added luxi call to submit multiple jobs in one go
1219 - Added cluster initialization option to not modify ``/etc/hosts``
1221 - Added network interface parameters
1222 - Added dry run mode to some LUs
1223 - Added RAPI resources:
1225 - ``/2/instances/[instance_name]/info``
1226 - ``/2/instances/[instance_name]/replace-disks``
1227 - ``/2/nodes/[node_name]/evacuate``
1228 - ``/2/nodes/[node_name]/migrate``
1229 - ``/2/nodes/[node_name]/role``
1230 - ``/2/nodes/[node_name]/storage``
1231 - ``/2/nodes/[node_name]/storage/modify``
1232 - ``/2/nodes/[node_name]/storage/repair``
1234 - Added OpCodes to evacuate or migrate all instances on a node
1235 - Added new command to list storage elements on nodes (``gnt-node
1236 list-storage``) and modify them (``gnt-node modify-storage``)
1237 - Added new ssconf files with master candidate IP address
1238 (``ssconf_master_candidates_ips``), node primary IP address
1239 (``ssconf_node_primary_ips``) and node secondary IP address
1240 (``ssconf_node_secondary_ips``)
1241 - Added ``ganeti-confd`` and a client library to query the Ganeti
1242 configuration via UDP
1243 - Added ability to run hooks after cluster initialization and before
1245 - Added automatic mode for disk replace (``gnt-instance replace-disks
1247 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1248 after catastrophic data-loss
1249 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1251 - Added ``gnt-instance move`` command to move instances
1252 - Added ``gnt-cluster watcher`` command to control watcher
1253 - Added ``gnt-node powercycle`` command to powercycle nodes
1254 - Added new job status field ``lock_status``
1255 - Added parseable error codes to cluster verification (``gnt-cluster
1256 verify --error-codes``) and made output less verbose (use
1257 ``--verbose`` to restore previous behaviour)
1258 - Added UUIDs to the main config entities (cluster, nodes, instances)
1259 - Added support for OS variants
1260 - Added support for hashed passwords in the Ganeti remote API users file
1262 - Added option to specify maximum timeout on instance shutdown
1263 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1264 - Added new helper script to start and stop Ganeti daemons
1265 (``daemon-util``), with the intent to reduce the work necessary to
1266 adjust Ganeti for non-Debian distributions and to start/stop daemons
1268 - Added more unittests
1269 - Fixed critical bug in ganeti-masterd startup
1270 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1271 now customisable at the cluster level for both the KVM and Xen
1272 hypervisors using the new ``migration_port`` parameter
1273 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1274 reinstalling an instance
1275 - Allowed ``@`` in tag names
1276 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1277 - Many documentation updates
1278 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1279 - ``gnt-instance reinstall`` can now reinstall multiple instances
1280 - Updated many command line parameters
1281 - Introduced new OS API version 15
1282 - No longer support a default hypervisor
1283 - Treat virtual LVs as inexistent
1284 - Improved job locking logic to reduce lock contention
1285 - Match instance and node names case insensitively
1286 - Reimplemented bash completion script to be more complete
1293 *(Released Thu, 4 Feb 2010)*
1295 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1296 - Fix a string formatting bug
1297 - Improve safety of the code in some error paths
1298 - Improve data validation in the master of values returned from nodes
1304 *(Released Thu, 17 Dec 2009)*
1306 - Fix security issue due to missing validation of iallocator names; this
1307 allows local and remote execution of arbitrary executables
1308 - Fix failure of gnt-node list during instance removal
1309 - Ship the RAPI documentation in the archive
1315 *(Released Wed, 30 Sep 2009)*
1317 - Fixed many wrong messages
1318 - Fixed a few bugs related to the locking library
1319 - Fixed MAC checking at instance creation time
1320 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1321 - Fixed a few issues related to signal handling in both daemons and
1323 - Fixed the example startup script provided
1324 - Fixed insserv dependencies in the example startup script (patch from
1326 - Fixed handling of drained nodes in the iallocator framework
1327 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1329 - Fixed error related to invalid job IDs in job polling
1330 - Fixed job/opcode persistence on unclean master shutdown
1331 - Fixed handling of partial job processing after unclean master
1333 - Fixed error reporting from LUs, previously all errors were converted
1334 into execution errors
1335 - Fixed error reporting from burnin
1336 - Decreased significantly the memory usage of the job queue
1337 - Optimised slightly multi-job submission
1338 - Optimised slightly opcode loading
1339 - Backported the multi-job submit framework from the development
1340 branch; multi-instance start and stop should be faster
1341 - Added script to clean archived jobs after 21 days; this will reduce
1342 the size of the queue directory
1343 - Added some extra checks in disk size tracking
1344 - Added an example ethers hook script
1345 - Added a cluster parameter that prevents Ganeti from modifying of
1347 - Added more node information to RAPI responses
1348 - Added a ``gnt-job watch`` command that allows following the ouput of a
1350 - Added a bind-address option to ganeti-rapi
1351 - Added more checks to the configuration verify
1352 - Enhanced the burnin script such that some operations can be retried
1354 - Converted instance reinstall to multi-instance model
1360 *(Released Fri, 7 Aug 2009)*
1362 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1363 to allow using the pre-2.0.2 behaviour in activation, if any existing
1364 instances have mismatched disk sizes in the configuration
1365 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1366 any configuration mismatches for disk sizes
1367 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1368 failover to work on two-node clusters
1369 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1371 - Fixed detection of OS script errors in ``gnt-backup export``
1372 - Fixed exit code of ``gnt-backup export``
1378 *(Released Fri, 17 Jul 2009)*
1380 - Added experimental support for stripped logical volumes; this should
1381 enhance performance but comes with a higher complexity in the block
1382 device handling; stripping is only enabled when passing
1383 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1384 affected even in the non-stripped mode
1385 - Improved resiliency against transient failures at the end of DRBD
1386 resyncs, and in general of DRBD resync checks
1387 - Fixed a couple of issues with exports and snapshot errors
1388 - Fixed a couple of issues in instance listing
1389 - Added display of the disk size in ``gnt-instance info``
1390 - Fixed checking for valid OSes in instance creation
1391 - Fixed handling of the "vcpus" parameter in instance listing and in
1392 general of invalid parameters
1393 - Fixed http server library, and thus RAPI, to handle invalid
1394 username/password combinations correctly; this means that now they
1395 report unauthorized for queries too, not only for modifications,
1396 allowing earlier detect of configuration problems
1397 - Added a new "role" node list field, equivalent to the master/master
1398 candidate/drained/offline flags combinations
1399 - Fixed cluster modify and changes of candidate pool size
1400 - Fixed cluster verify error messages for wrong files on regular nodes
1401 - Fixed a couple of issues with node demotion from master candidate role
1402 - Fixed node readd issues
1403 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1404 - Added a new ``--no-voting`` option for masterfailover to fix failover
1405 on two-nodes clusters when the former master node is unreachable
1406 - Added instance reinstall over RAPI
1412 *(Released Tue, 16 Jun 2009)*
1414 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1415 allow re-adding the start in single-user option (regression from 1.2)
1416 - the watcher writes the instance status to a file, to allow monitoring
1417 to report the instance status (from the master) based on cached
1418 results of the watcher's queries; while this can get stale if the
1419 watcher is being locked due to other work on the cluster, this is
1420 still an improvement
1421 - the watcher now also restarts the node daemon and the rapi daemon if
1423 - fixed the watcher to handle full and drained queue cases
1424 - hooks export more instance data in the environment, which helps if
1425 hook scripts need to take action based on the instance's properties
1426 (no longer need to query back into ganeti)
1427 - instance failovers when the instance is stopped do not check for free
1428 RAM, so that failing over a stopped instance is possible in low memory
1430 - rapi uses queries for tags instead of jobs (for less job traffic), and
1431 for cluster tags it won't talk to masterd at all but read them from
1433 - a couple of error handling fixes in RAPI
1434 - drbd handling: improved the error handling of inconsistent disks after
1435 resync to reduce the frequency of "there are some degraded disks for
1436 this instance" messages
1437 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1438 error handling path called a wrong function name)
1444 *(Released Wed, 27 May 2009)*
1446 - no changes from rc5
1449 Version 2.0 release candidate 5
1450 -------------------------------
1452 *(Released Wed, 20 May 2009)*
1454 - fix a couple of bugs (validation, argument checks)
1455 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1456 - some small improvements to RAPI and IAllocator
1457 - make watcher automatically start the master daemon if down
1460 Version 2.0 release candidate 4
1461 -------------------------------
1463 *(Released Mon, 27 Apr 2009)*
1465 - change the OS list to not require locks; this helps with big clusters
1466 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1467 volume group is broken
1468 - ``gnt-instance info``, without any arguments, doesn't run for all
1469 instances anymore; either pass ``--all`` or pass the desired
1470 instances; this helps against mistakes on big clusters where listing
1471 the information for all instances takes a long time
1472 - miscellaneous doc and man pages fixes
1475 Version 2.0 release candidate 3
1476 -------------------------------
1478 *(Released Wed, 8 Apr 2009)*
1480 - Change the internal locking model of some ``gnt-node`` commands, in
1481 order to reduce contention (and blocking of master daemon) when
1482 batching many creation/reinstall jobs
1483 - Fixes to Xen soft reboot
1484 - No longer build documentation at build time, instead distribute it in
1485 the archive, in order to reduce the need for the whole docbook/rst
1489 Version 2.0 release candidate 2
1490 -------------------------------
1492 *(Released Fri, 27 Mar 2009)*
1494 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1495 - Fix watcher startup sequence, improves the behaviour of busy clusters
1496 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1497 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1498 bind address and other places
1499 - Some documentation fixes and updates
1502 Version 2.0 release candidate 1
1503 -------------------------------
1505 *(Released Mon, 2 Mar 2009)*
1507 - More documentation updates, now all docs should be more-or-less
1509 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1511 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1513 - Fix an issue related to $libdir/run/ganeti and cluster creation
1519 *(Released Thu, 19 Feb 2009)*
1521 - Xen PVM and KVM have switched the default value for the instance root
1522 disk to the first partition on the first drive, instead of the whole
1523 drive; this means that the OS installation scripts must be changed
1525 - Man pages have been updated
1526 - RAPI has been switched by default to HTTPS, and the exported functions
1527 should all work correctly
1528 - RAPI v1 has been removed
1529 - Many improvements to the KVM hypervisor
1530 - Block device errors are now better reported
1531 - Many other bugfixes and small improvements
1537 *(Released Mon, 26 Jan 2009)*
1539 - Version 2 is a general rewrite of the code and therefore the
1540 differences are too many to list, see the design document for 2.0 in
1541 the ``doc/`` subdirectory for more details
1542 - In this beta version there is not yet a migration path from 1.2 (there
1543 will be one in the final 2.0 release)
1544 - A few significant changes are:
1546 - all commands are executed by a daemon (``ganeti-masterd``) and the
1547 various ``gnt-*`` commands are just front-ends to it
1548 - all the commands are entered into, and executed from a job queue,
1549 see the ``gnt-job(8)`` manpage
1550 - the RAPI daemon supports read-write operations, secured by basic
1551 HTTP authentication on top of HTTPS
1552 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1553 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1554 to migrate to DRBD 8 first while still running Ganeti 1.2)
1555 - DRBD devices are using statically allocated minor numbers, which
1556 will be assigned to existing instances during the migration process
1557 - there is support for both Xen PVM and Xen HVM instances running on
1559 - KVM virtualization is supported too
1560 - file-based storage has been implemented, which means that it is
1561 possible to run the cluster without LVM and DRBD storage, for
1562 example using a shared filesystem exported from shared storage (and
1563 still have live migration)
1569 *(Released Tue, 13 Jan 2009)*
1571 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1572 - Reuse the old instance mac address by default on instance import, if
1573 the instance name is the same.
1574 - Handle situations in which the node info rpc returns incomplete
1576 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1577 - Improved version of batcher:
1579 - state file support
1580 - instance mac address support
1581 - support for HVM clusters/instances
1583 - Add an option to show the number of cpu sockets and nodes in
1585 - Support OSes that handle more than one version of the OS api (but do
1586 not change the current API in any other way)
1587 - Fix ``gnt-node migrate``
1588 - ``gnt-debug`` man page
1589 - Fixes various more typos and small issues
1590 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1596 *(Released Wed, 24 Sep 2008)*
1598 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1599 type of disk exported to fully virtualized instances.
1600 - provide access to the serial console of HVM instances
1601 - instance auto_balance flag, set by default. If turned off it will
1602 avoid warnings on cluster verify if there is not enough memory to fail
1603 over an instance. in the future it will prevent automatically failing
1604 it over when we will support that.
1605 - batcher tool for instance creation, see ``tools/README.batcher``
1606 - ``gnt-instance reinstall --select-os`` to interactively select a new
1607 operating system when reinstalling an instance.
1608 - when changing the memory amount on instance modify a check has been
1609 added that the instance will be able to start. also warnings are
1610 emitted if the instance will not be able to fail over, if auto_balance
1612 - documentation fixes
1613 - sync fields between ``gnt-instance list/modify/add/import``
1614 - fix a race condition in drbd when the sync speed was set after giving
1615 the device a remote peer.
1621 *(Released Tue, 22 Jul 2008)*
1623 - note: the allowed size and number of tags per object were reduced
1624 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1625 - fixed twisted 8.x compatibility
1626 - fixed ``gnt-instance replace-disks`` with iallocator
1627 - add TCP keepalives on twisted connections to detect restarted nodes
1628 - disk increase support, see ``gnt-instance grow-disk``
1629 - implement bulk node/instance query for RAPI
1630 - add tags in node/instance listing (optional)
1631 - experimental migration (and live migration) support, read the man page
1632 for ``gnt-instance migrate``
1633 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1634 has some small improvements in handling its state file
1640 *(Released Fri, 13 Jun 2008)*
1642 - Experimental readonly, REST-based remote API implementation;
1643 automatically started on master node, TCP port 5080, if enabled by
1644 ``--enable-rapi`` parameter to configure script.
1645 - Instance allocator support. Add and import instance accept a
1646 ``--iallocator`` parameter, and call that instance allocator to decide
1647 which node to use for the instance. The iallocator document describes
1648 what's expected from an allocator script.
1649 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1650 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1651 node is lost there is still enough memory to fail over the instances
1653 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1654 ``gnt-cluster verify``, to check for site-specific compliance. All the
1655 hooks will run, and their output, if any, will be displayed. Any
1656 failing hook will make the verification return an error value.
1657 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1658 primary and secondary interfaces
1659 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1660 that is still declared as part of the cluster and has failed.
1661 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1662 specifying output fields, that just adds the chosen fields to the
1664 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1665 export from the filesystem.
1666 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1667 have been added. Using them you can enable/disable acpi and pae
1668 support, and specify a path for a cd image to be exported to the
1669 instance. These parameters as the name suggest only work on HVM
1671 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1672 PAE support will be set to the previously hardcoded values, but the
1673 (previously hardcoded) path to the CDROM ISO image will be unset and
1674 if required, needs to be set manually with ``gnt-instance modify``
1676 - The address to which an instance's VNC console is bound is now
1677 selectable per-instance, rather than being cluster wide. Of course
1678 this only applies to instances controlled via VNC, so currently just
1679 applies to HVM clusters.
1685 *(Released Mon, 18 Feb 2008)*
1687 - more tweaks to the disk activation code (especially helpful for DRBD)
1688 - change the default ``gnt-instance list`` output format, now there is
1689 one combined status field (see the manpage for the exact values this
1691 - some more fixes for the mac export to hooks change
1692 - make Ganeti not break with DRBD 8.2.x (which changed the version
1693 format in ``/proc/drbd``) (issue 24)
1694 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1695 template, allowing migration from DRBD0.7+MD to DRBD8
1701 *(Released Wed, 30 Jan 2008)*
1703 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1705 - add command aliases infrastructure and a few aliases
1706 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1707 man pages and the ``--help`` option of ``gnt-node
1708 list``/``gnt-instance list``
1709 - fix ``gnt-backup list`` with down nodes (issue 21)
1710 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1711 - fix the dist archive and add a check for including svn/git files in
1713 - some developer-related changes: improve the burnin and the QA suite,
1714 add an upload script for testing during development
1720 *(Released Wed, 16 Jan 2008)*
1722 - experimental HVM support, read the install document, section
1723 "Initializing the cluster"
1724 - allow for the PVM hypervisor per-instance kernel and initrd paths
1725 - add a new command ``gnt-cluster verify-disks`` which uses a new
1726 algorithm to improve the reconnection of the DRBD pairs if the device
1727 on the secondary node has gone away
1728 - make logical volume code auto-activate LVs at disk activation time
1729 - slightly improve the speed of activating disks
1730 - allow specification of the MAC address at instance creation time, and
1731 changing it later via ``gnt-instance modify``
1732 - fix handling of external commands that generate lots of output on
1734 - update documentation with regard to minimum version of DRBD8 supported
1740 *(Released Tue, 4 Dec 2007)*
1742 - Log the ``xm create`` output to the node daemon log on failure (to
1743 help diagnosing the error)
1744 - In debug mode, log all external commands output if failed to the logs
1745 - Change parsing of lvm commands to ignore stderr
1751 *(Released Wed, 28 Nov 2007)*
1753 - Another round of updates to the DRBD 8 code to deal with more failures
1754 in the replace secondary node operation
1755 - Some more logging of failures in disk operations (lvm, drbd)
1756 - A few documentation updates
1763 *(Released Tue, 13 Nov 2007)*
1765 - Change configuration file format from Python's Pickle to JSON.
1766 Upgrading is possible using the cfgupgrade utility.
1767 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1768 faster replace disks and is more stable (DRBD 8 has many improvements
1769 compared to DRBD 0.7)
1770 - Added command line tags support (see man pages for ``gnt-instance``,
1771 ``gnt-node``, ``gnt-cluster``)
1772 - Added instance rename support
1773 - Added multi-instance startup/shutdown
1774 - Added cluster rename support
1775 - Added ``gnt-node evacuate`` to simplify some node operations
1776 - Added instance reboot operation that can speedup reboot as compared to
1778 - Soften the requirement that hostnames are in FQDN format
1779 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1781 - Removed dependency on debian's patched fping that uses the
1782 non-standard ``-S`` option
1783 - Now the OS definitions are searched for in multiple, configurable
1784 paths (easier for distros to package)
1785 - Some changes to the hooks infrastructure (especially the new
1786 post-configuration update hook)
1787 - Other small bugfixes
1789 .. vim: set textwidth=72 syntax=rst :