10 - Deprecated ``admin_up`` field. Instead, ``admin_state`` is introduced,
11 with 3 possible values -- ``up``, ``down`` and ``offline``.
19 Improvements and bugfixes
20 ~~~~~~~~~~~~~~~~~~~~~~~~~
22 - Support for kvm version 1.0, that changed the version reporting format
29 *(Released Thu, 27 Oct 2011)*
31 Incompatible/important changes and bugfixes
32 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
34 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
35 resource's ``ip_check`` parameter changed from ``True`` to ``False``
36 to match the underlying LUXI interface.
37 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
38 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
39 not maintain backwards-compatibility as the underlying operation
40 changed in an incompatible way. The RAPI client can talk to old
41 servers, but it needs to be told so as the return value changed.
42 - When creating file-based instances via RAPI, the ``file_driver``
43 parameter no longer defaults to ``loop`` and must be specified.
44 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
46 - Support for the undocumented and deprecated RAPI instance creation
47 request format version 0 has been dropped. Use version 1, supported
48 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
49 - Pyparsing 1.4.6 or above is required, see :doc:`installation
50 documentation <install>`.
51 - The "cluster-verify" hooks are now executed per group by the
52 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
53 if you just run ``gnt-cluster verify``, which generates one opcode per
55 - The environment as passed to the OS scripts is cleared, and thus no
56 environment variables defined in the node daemon's environment will be
57 inherited by the scripts.
58 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
60 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
61 support operations involving multiple node groups.
62 - Offline nodes are ignored when failing over an instance.
67 - The ganeti-htools project has been merged into the ganeti-core source
68 tree and will be built as part of Ganeti (see :doc:`install-quick`).
69 - Implemented support for :doc:`shared storage <design-shared-storage>`.
70 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
71 GPT-style partition tables (requires `parted
72 <http://www.gnu.org/s/parted/>`_).
73 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
74 - Allowed adding tags on instance creation.
75 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
76 - Allow instances to be started in a paused state, enabling the user to
77 see the complete console output on boot using the console.
78 - Added new hypervisor flag to control default reboot behaviour
79 (``reboot_behavior``).
80 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
81 - Improved out-of-band management support:
83 - Added ``gnt-node health`` command reporting the health status of
85 - Added ``gnt-node power`` command to manage power status of nodes.
86 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
88 - Instance migration can fall back to failover if instance is not
90 - Filters can be used when listing nodes, instances, groups and locks;
91 see *ganeti(7)* manpage.
92 - Added post-execution status as variables to :doc:`hooks <hooks>`
94 - Instance tags are exported/imported together with the instance.
95 - When given an explicit job ID, ``gnt-job info`` will work for archived
97 - Jobs can define dependencies on other jobs (not yet supported via
98 RAPI or command line, but used by internal commands and usable via
101 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
104 - Instance failover is now available as a RAPI resource
105 (``/2/instances/[instance_name]/failover``).
106 - ``gnt-instance info`` defaults to static information if primary node
108 - Opcodes have a new ``comment`` attribute.
109 - Added basic SPICE support to KVM hypervisor.
110 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
112 Node group improvements
113 ~~~~~~~~~~~~~~~~~~~~~~~
115 - ``gnt-cluster verify`` has been modified to check groups separately,
116 thereby improving performance.
117 - Node group support has been added to ``gnt-cluster verify-disks``,
118 which now operates per node group.
119 - Watcher has been changed to work better with node groups.
121 - One process and state file per node group.
122 - Slow watcher in one group doesn't block other group's watcher.
124 - Added new command, ``gnt-group evacuate``, to move all instances in a
125 node group to other groups.
126 - Added ``gnt-instance change-group`` to move an instance to another
128 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
129 per-group operations.
130 - Node groups can be tagged.
131 - Some operations switch from an exclusive to a shared lock as soon as
133 - Instance's primary and secondary nodes' groups are now available as
134 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
135 and ``snodes.group.uuid``).
140 - Numerous updates to documentation and manpages.
142 - :doc:`RAPI <rapi>` documentation now has detailed parameter
144 - Some opcode/job results are now also documented, see :doc:`RAPI
147 - A lockset's internal lock is now also visible in lock monitor.
148 - Log messages from job queue workers now contain information about the
149 opcode they're processing.
150 - ``gnt-instance console`` no longer requires the instance lock.
151 - A short delay when waiting for job changes reduces the number of LUXI
152 requests significantly.
153 - DRBD metadata volumes are overwritten with zeros during disk creation.
159 *(Released Wed, 26 Oct 2011)*
161 This was the third release candidate of the 2.5 series.
167 *(Released Tue, 18 Oct 2011)*
169 This was the second release candidate of the 2.5 series.
175 *(Released Tue, 4 Oct 2011)*
177 This was the first release candidate of the 2.5 series.
183 *(Released Wed, 31 Aug 2011)*
185 This was the third beta release of the 2.5 series.
191 *(Released Mon, 22 Aug 2011)*
193 This was the second beta release of the 2.5 series.
199 *(Released Fri, 12 Aug 2011)*
201 This was the first beta release of the 2.5 series.
207 *(Released Thu, 27 Oct 2011)*
209 - Fixed bug when parsing command line parameter values ending in
211 - Fixed assertion error after unclean master shutdown
212 - Disable HTTP client pool for RPC, significantly reducing memory usage
214 - Fixed queue archive creation with wrong permissions
220 *(Released Tue, 23 Aug 2011)*
224 - Fixed documentation for importing with ``--src-dir`` option
225 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
226 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
232 *(Released Fri, 5 Aug 2011)*
234 Many bug-fixes and a few small features:
236 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
237 issues when you tried to add an instance with two MAC addresses in one
239 - KVM: fixed per-instance stored UID value
240 - KVM: configure bridged NICs at migration start
241 - KVM: Fix a bug where instance will not start with never KVM versions
243 - Added OS search path to ``gnt-cluster info``
244 - Fixed an issue with ``file_storage_dir`` where you were forced to
245 provide an absolute path, but the documentation states it is a
246 relative path, the documentation was right
247 - Added a new parameter to instance stop/start called ``--no-remember``
248 that will make the state change to not be remembered
249 - Implemented ``no_remember`` at RAPI level
250 - Improved the documentation
251 - Node evacuation: don't call IAllocator if node is already empty
252 - Fixed bug in DRBD8 replace disks on current nodes
253 - Fixed bug in recreate-disks for DRBD instances
254 - Moved assertion checking locks in ``gnt-instance replace-disks``
255 causing it to abort with not owning the right locks for some situation
256 - Job queue: Fixed potential race condition when cancelling queued jobs
257 - Fixed off-by-one bug in job serial generation
258 - ``gnt-node volumes``: Fix instance names
259 - Fixed aliases in bash completion
260 - Fixed a bug in reopening log files after being sent a SIGHUP
261 - Added a flag to burnin to allow specifying VCPU count
262 - Bugfixes to non-root Ganeti configuration
268 *(Released Thu, 12 May 2011)*
270 Many bug-fixes and a few new small features:
272 - Fixed a bug related to log opening failures
273 - Fixed a bug in instance listing with orphan instances
274 - Fixed a bug which prevented resetting the cluster-level node parameter
275 ``oob_program`` to the default
276 - Many fixes related to the ``cluster-merge`` tool
277 - Fixed a race condition in the lock monitor, which caused failures
278 during (at least) creation of many instances in parallel
279 - Improved output for gnt-job info
280 - Removed the quiet flag on some ssh calls which prevented debugging
282 - Improved the N+1 failure messages in cluster verify by actually
283 showing the memory values (needed and available)
284 - Increased lock attempt timeouts so that when executing long operations
285 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
286 too early and thus prevent the use of the 'fair' mechanism
287 - Changed instance query data (``gnt-instance info``) to not acquire
288 locks unless needed, thus allowing its use on locked instance if only
289 static information is asked for
290 - Improved behaviour with filesystems that do not support rename on an
292 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
293 kept locks on all nodes during the wipe, which is unneeded
294 - Fixed ``gnt-watcher`` handling of errors during hooks execution
295 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
296 10GiB) which caused the wipe to fail right at the end in some cases
297 - Fixed master IP activation when doing master failover with no-voting
298 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
299 the master node itself
300 - Fixed potential data-loss in under disk full conditions, where Ganeti
301 wouldn't check correctly the return code and would consider
302 partially-written files 'correct'
303 - Fixed bug related to multiple VGs and DRBD disk replacing
304 - Added new disk parameter ``metavg`` that allows placement of the meta
305 device for DRBD in a different volume group
306 - Fixed error handling in the node daemon when the system libc doesn't
307 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
308 - Fixed lock release during replace-disks, which kept cluster-wide locks
309 when doing disk replaces with an iallocator script
310 - Added check for missing bridges in cluster verify
311 - Handle EPIPE errors while writing to the terminal better, so that
312 piping the output to e.g. ``less`` doesn't cause a backtrace
313 - Fixed rare case where a ^C during Luxi calls could have been
314 interpreted as server errors, instead of simply terminating
315 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
317 - Added a few more parameters to the KVM hypervisor, allowing a second
318 CDROM, custom disk type for CDROMs and a floppy image
319 - Removed redundant message in instance rename when the name is given
321 - Added option to ``gnt-instance recreate-disks`` to allow creating the
322 disks on new nodes, allowing recreation when the original instance
323 nodes are completely gone
324 - Added option when converting disk templates to DRBD to skip waiting
325 for the resync, in order to make the instance available sooner
326 - Added two new variables to the OS scripts environment (containing the
328 - Made the root_path and optional parameter for the xen-pvm hypervisor,
329 to allow use of ``pvgrub`` as bootloader
330 - Changed the instance memory modifications to only check out-of-memory
331 conditions on memory increases, and turned the secondary node warnings
332 into errors (they can still be overridden via ``--force``)
333 - Fixed the handling of a corner case when the Python installation gets
334 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
335 to execute a command that doesn't exist
336 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
337 node of the instance returned failures during instance shutdown; this
338 adds the option ``--ignore-consistency`` to gnt-instance move
340 And as usual, various improvements to the error messages, documentation
347 *(Released Wed, 09 Mar 2011)*
349 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
350 the RAPI users file if run twice (even with ``--dry-run``).
352 The release fixes that bug (nothing else changed).
358 *(Released Mon, 07 Mar 2011)*
360 Final 2.4.0 release. Just a few small fixes:
362 - Fixed RAPI node evacuate
363 - Fixed the kvm-ifup script
364 - Fixed internal error handling for special job cases
365 - Updated man page to specify the escaping feature for options
371 *(Released Mon, 28 Feb 2011)*
373 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
374 that this feature wiped the disks of the wrong instance, leading to loss
379 - Fixed title of query field containing instance name
380 - Expanded the glossary in the documentation
381 - Fixed one unittest (internal issue)
387 *(Released Mon, 21 Feb 2011)*
389 A number of bug fixes plus just a couple functionality changes.
391 On the user-visible side, the ``gnt-* list`` command output has changed
392 with respect to "special" field states. The current rc1 style of display
393 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
394 the default output mode special fields states are displayed as follows:
396 - Offline resource: ``*``
397 - Unavailable/not applicable: ``-``
398 - Data missing (RPC failure): ``?``
399 - Unknown field: ``??``
401 Another user-visible change is the addition of ``--force-join`` to
406 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
407 - Fixed regression in RAPI/instance reinstall where all parameters were
408 required (instead of optional)
409 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
410 - Fixed iallocator usage (offline nodes were not considered offline)
411 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
412 - Fixed hypervisor and OS parameter validation with respect to
414 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
416 - Fixed ``tools/listrunner`` with respect to agent-based usage
422 *(Released Fri, 4 Feb 2011)*
424 Many changes and fixes since the beta1 release. While there were some
425 internal changes, the code has been mostly stabilised for the RC
428 Note: the dumb allocator was removed in this release, as it was not kept
429 up-to-date with the IAllocator protocol changes. It is recommended to
430 use the ``hail`` command from the ganeti-htools package.
432 Note: the 2.4 and up versions of Ganeti are not compatible with the
433 0.2.x branch of ganeti-htools. You need to upgrade to
434 ganeti-htools-0.3.0 (or later).
436 Regressions fixed from 2.3
437 ~~~~~~~~~~~~~~~~~~~~~~~~~~
439 - Fixed the ``gnt-cluster verify-disks`` command
440 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
442 - Fixed disk adoption breakage
443 - Fixed wrong headers in instance listing for field aliases
448 - Fixed corner case in KVM handling of NICs
449 - Fixed many cases of wrong handling of non-vm_capable nodes
450 - Fixed a bug where a missing instance symlink was not possible to
451 recreate with any ``gnt-*`` command (now ``gnt-instance
452 activate-disks`` does it)
453 - Fixed the volume group name as reported by ``gnt-cluster
455 - Increased timeouts for the import-export code, hopefully leading to
456 fewer aborts due network or instance timeouts
457 - Fixed bug in ``gnt-node list-storage``
458 - Fixed bug where not all daemons were started on cluster
459 initialisation, but only at the first watcher run
460 - Fixed many bugs in the OOB implementation
461 - Fixed watcher behaviour in presence of instances with offline
463 - Fixed instance list output for instances running on the wrong node
464 - a few fixes to the cluster-merge tool, but it still cannot merge
465 multi-node groups (currently it is not recommended to use this tool)
471 - Improved network configuration for the KVM hypervisor
472 - Added e1000 as a supported NIC for Xen-HVM
473 - Improved the lvmstrap tool to also be able to use partitions, as
474 opposed to full disks
475 - Improved speed of disk wiping (the cluster parameter
476 ``prealloc_wipe_disks``, so that it has a low impact on the total time
477 of instance creations
478 - Added documentation for the OS parameters
479 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
480 hypervisor is not responding
481 - Added display of blacklisted and hidden OS information in
483 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
484 NIC and node parameters, which might create problems with currently
485 invalid (but undetected) configuration files, but prevents validation
486 failures when unrelated parameters are modified
487 - Changed cluster initialisation to wait for the master daemon to become
489 - Expanded the RAPI interface:
491 - Added config redistribution resource
492 - Added activation/deactivation of instance disks
493 - Added export of console information
495 - Implemented log file reopening on SIGHUP, which allows using
496 logrotate(8) for the Ganeti log files
497 - Added a basic OOB helper script as an example
503 *(Released Fri, 14 Jan 2011)*
508 - Fixed timezone issues when formatting timestamps
509 - Added support for node groups, available via ``gnt-group`` and other
511 - Added out-of-band framework and management, see :doc:`design
512 document <design-oob>`
513 - Removed support for roman numbers from ``gnt-node list`` and
514 ``gnt-instance list``.
515 - Allowed modification of master network interface via ``gnt-cluster
516 modify --master-netdev``
517 - Accept offline secondaries while shutting down instance disks
518 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
519 - Added support for multiple LVM volume groups
520 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
522 - Added commands to list available fields:
524 - ``gnt-node list-fields``
525 - ``gnt-group list-fields``
526 - ``gnt-instance list-fields``
528 - Updated documentation and man pages
533 - Moved ``rapi_users`` file into separate directory, now named
534 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
536 - Added new tool for running commands on many machines,
537 ``tools/ganeti-listrunner``
538 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
539 improving the ``gnt-instance console`` output
540 - Allowed customisation of disk index separator at ``configure`` time
541 - Export node group allocation policy to :doc:`iallocator <iallocator>`
542 - Added support for non-partitioned md disks in ``lvmstrap``
543 - Added script to gracefully power off KVM instances
544 - Split ``utils`` module into smaller parts
545 - Changed query operations to return more detailed information, e.g.
546 whether an information is unavailable due to an offline node. To use
547 this new functionality, the LUXI call ``Query`` must be used. Field
548 information is now stored by the master daemon and can be retrieved
549 using ``QueryFields``. Instances, nodes and groups can also be queried
550 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
551 exposed via RAPI). The following commands make use of this
552 infrastructure change:
555 - ``gnt-group list-fields``
557 - ``gnt-node list-fields``
558 - ``gnt-instance list``
559 - ``gnt-instance list-fields``
560 - ``gnt-debug locks``
565 - New RAPI resources (see :doc:`rapi`):
569 - ``/2/groups/[group_name]``
570 - ``/2/groups/[group_name]/assign-nodes``
571 - ``/2/groups/[group_name]/modify``
572 - ``/2/groups/[group_name]/rename``
573 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
577 - Implemented ``no_install`` for instance creation
578 - Implemented OS parameters for instance reinstallation, allowing
579 use of special settings on reinstallation (e.g. for preserving data)
584 - Added IPv6 support in import/export
585 - Pause DRBD synchronization while wiping disks on instance creation
586 - Updated unittests and QA scripts
587 - Improved network parameters passed to KVM
588 - Converted man pages from docbook to reStructuredText
594 *(Released Mon, 20 Dec 2010)*
596 Released version 2.3.1~rc1 without any changes.
602 *(Released Wed, 1 Dec 2010)*
604 - impexpd: Disable OpenSSL compression in socat if possible (backport
605 from master, commit e90739d625b, see :doc:`installation guide
606 <install-quick>` for details)
607 - Changed unittest coverage report to exclude test scripts
608 - Added script to check version format
614 *(Released Wed, 1 Dec 2010)*
616 Released version 2.3.0~rc1 without any changes.
622 *(Released Fri, 19 Nov 2010)*
624 A number of bugfixes and documentation updates:
626 - Update ganeti-os-interface documentation
627 - Fixed a bug related to duplicate MACs or similar items which should be
629 - Fix breakage in OS state modify
630 - Reinstall instance: disallow offline secondaries (fixes bug related to
631 OS changing but reinstall failing)
632 - plus all the other fixes between 2.2.1 and 2.2.2
638 *(Released Tue, 2 Nov 2010)*
640 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
641 - Fixed master failover race with watcher
642 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
644 - Accept previously stopped instance for export with instance removal
645 - Simplify and extend the environment variables for instance OS scripts
646 - Added new node flags, ``master_capable`` and ``vm_capable``
647 - Added optional instance disk wiping prior during allocation. This is a
648 cluster-wide option and can be set/modified using
649 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
650 - Added IPv6 support, see :doc:`design document <design-2.3>` and
652 - Added a new watcher option (``--ignore-pause``)
653 - Added option to ignore offline node on instance start/stop
654 (``--ignore-offline``)
655 - Allow overriding OS parameters with ``gnt-instance reinstall``
656 - Added ability to change node's secondary IP address using ``gnt-node
658 - Implemented privilege separation for all daemons except
659 ``ganeti-noded``, see ``configure`` options
660 - Complain if an instance's disk is marked faulty in ``gnt-cluster
662 - Implemented job priorities (see ``ganeti(7)`` manpage)
663 - Ignore failures while shutting down instances during failover from
665 - Exit daemon's bootstrap process only once daemon is ready
666 - Export more information via ``LUInstanceQuery``/remote API
667 - Improved documentation, QA and unittests
668 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
669 restart if the file was created or changed
670 - Added LUXI protocol version sent with each request and response,
671 allowing detection of server/client mismatches
672 - Moved the Python scripts among gnt-* and ganeti-* into modules
673 - Moved all code related to setting up SSH to an external script,
675 - Infrastructure changes for node group support in future versions
681 *(Released Fri, 19 Nov 2010)*
683 A few small bugs fixed, and some improvements to the build system:
685 - Fix documentation regarding conversion to drbd
686 - Fix validation of parameters in cluster modify (``gnt-cluster modify
688 - Fix error handling in node modify with multiple changes
689 - Allow remote imports without checked names
695 *(Released Tue, 19 Oct 2010)*
697 - Disable SSL session ID cache in RPC client
703 *(Released Thu, 14 Oct 2010)*
705 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
706 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
708 - Fix problems with interaction between Curl and Python's HTTP server,
709 resulting in increased speed in many RPC calls
710 - Improve our release script to prevent breakage with older aclocal and
717 *(Released Thu, 7 Oct 2010)*
719 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
720 value retrieved from master
721 - Added support for blacklisted or hidden OS definitions
722 - Added simple lock monitor (accessible via (``gnt-debug locks``)
723 - Added support for -mem-path in KVM hypervisor abstraction layer
724 - Allow overriding instance parameters in tool for inter-cluster
725 instance moves (``tools/move-instance``)
726 - Improved opcode summaries (e.g. in ``gnt-job list``)
727 - Improve consistency of OS listing by sorting it
728 - Documentation updates
734 *(Released Fri, 8 Oct 2010)*
736 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
742 *(Released Mon, 4 Oct 2010)*
744 - Fixed regression in ``gnt-instance rename``
750 *(Released Wed, 22 Sep 2010)*
752 - Fixed OS_VARIANT variable for OS scripts
753 - Fixed cluster tag operations via RAPI
754 - Made ``setup-ssh`` exit with non-zero code if an error occurred
755 - Disabled RAPI CA checks in watcher
761 *(Released Mon, 23 Aug 2010)*
763 - Support DRBD versions of the format "a.b.c.d"
765 - Re-introduce support for usage from multiple threads in RAPI client
766 - Instance renames and modify via RAPI
767 - Work around race condition between processing and archival in job
769 - Mark opcodes following failed one as failed, too
770 - Job field ``lock_status`` was removed due to difficulties making it
771 work with the changed job queue in Ganeti 2.2; a better way to monitor
772 locks is expected for a later 2.2.x release
773 - Fixed dry-run behaviour with many commands
774 - Support ``ssh-agent`` again when adding nodes
775 - Many additional bugfixes
781 *(Released Fri, 30 Jul 2010)*
783 Important change: the internal RPC mechanism between Ganeti nodes has
784 changed from using a home-grown http library (based on the Python base
785 libraries) to use the PycURL library. This requires that PycURL is
786 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
787 against GnuTLS by default. cURL's support for GnuTLS had known issues
788 before cURL 7.21.0 and we recommend using the latest cURL release or
789 linking against OpenSSL. Most other distributions already link PycURL
790 and cURL against OpenSSL. The command::
792 python -c 'import pycurl; print pycurl.version'
794 can be used to determine the libraries PycURL and cURL are linked
797 Other significant changes:
799 - Rewrote much of the internals of the job queue, in order to achieve
800 better parallelism; this decouples job query operations from the job
801 processing, and it should allow much nicer behaviour of the master
802 daemon under load, and it also has uncovered some long-standing bugs
803 related to the job serialisation (now fixed)
804 - Added a default iallocator setting to the cluster parameters,
805 eliminating the need to always pass nodes or an iallocator for
806 operations that require selection of new node(s)
807 - Added experimental support for the LXC virtualization method
808 - Added support for OS parameters, which allows the installation of
809 instances to pass parameter to OS scripts in order to customise the
811 - Added a hypervisor parameter controlling the migration type (live or
812 non-live), since hypervisors have various levels of reliability; this
813 has renamed the 'live' parameter to 'mode'
814 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
815 logical volumes, meaning that cluster verify will ignore them and not
816 flag their presence as errors
817 - The watcher will now reset the error count for failed instances after
818 8 hours, thus allowing self-healing if the problem that caused the
819 instances to be down/fail to start has cleared in the meantime
820 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
821 check for, and warn, if the drbd module parameter ``usermode_helper``
822 is not consistent with the cluster-wide setting; this is needed to
823 make diagnose easier of failed drbd creations
824 - Started adding base IPv6 support, but this is not yet
825 enabled/available for use
826 - Rename operations (cluster, instance) will now return the new name,
827 which is especially useful if a short name was passed in
828 - Added support for instance migration in RAPI
829 - Added a tool to pre-configure nodes for the SSH setup, before joining
830 them to the cluster; this will allow in the future a simplified model
831 for node joining (but not yet fully enabled in 2.2); this needs the
832 paramiko python library
833 - Fixed handling of name-resolving errors
834 - Fixed consistency of job results on the error path
835 - Fixed master-failover race condition when executed multiple times in
837 - Fixed many bugs related to the job queue (mostly introduced during the
838 2.2 development cycle, so not all are impacting 2.1)
839 - Fixed instance migration with missing disk symlinks
840 - Fixed handling of unknown jobs in ``gnt-job archive``
841 - And many other small fixes/improvements
845 - Enhanced both the unittest and the QA coverage
846 - Switched the opcode validation to a generic model, and extended the
847 validation to all opcode parameters
848 - Changed more parts of the code that write shell scripts to use the
850 - Switched the master daemon to use the asyncore library for the Luxi
857 *(Released Thu, 17 Jun 2010)*
859 - Added tool (``move-instance``) and infrastructure to move instances
860 between separate clusters (see :doc:`separate documentation
861 <move-instance>` and :doc:`design document <design-2.2>`)
862 - Added per-request RPC timeout
863 - RAPI now requires a Content-Type header for requests with a body (e.g.
864 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
865 :rfc:`2616` (HTTP/1.1), section 7.2.1)
866 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
868 - Implemented initial support for running Ganeti daemons as separate
869 users, see configure-time flags ``--with-user-prefix`` and
870 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
872 - Instances can be removed after export (``gnt-backup export
874 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
875 key (instead of 1024 bit)
876 - Added new cluster configuration file for cluster domain secret
877 - Import/export now use SSL instead of SSH
878 - Added support for showing estimated time when exporting an instance,
879 see the ``ganeti-os-interface(7)`` manpage and look for
886 *(Released Tue, 16 Nov 2010)*
888 Some more bugfixes. Unless critical bugs occur, this will be the last
891 - Fix case of MAC special-values
892 - Fix mac checker regex
893 - backend: Fix typo causing "out of range" error
894 - Add missing --units in gnt-instance list man page
900 *(Released Tue, 24 Aug 2010)*
903 - Don't ignore secondary node silently on non-mirrored disk templates
905 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
906 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
907 - Properly document the usb_mouse parameter
908 - Fix path in ganeti-rapi(8) (issue 116)
909 - Adjust error message when the ganeti user's .ssh directory is
911 - Add same-node-check when changing the disk template to drbd
917 *(Released Fri, 16 Jul 2010)*
920 - Add an option to only select some reboot types during qa/burnin.
921 (on some hypervisors consequent reboots are not supported)
922 - Fix infrequent race condition in master failover. Sometimes the old
923 master ip address would be still detected as up for a short time
924 after it was removed, causing failover to fail.
925 - Decrease mlockall warnings when the ctypes module is missing. On
926 Python 2.4 we support running even if no ctypes module is installed,
927 but we were too verbose about this issue.
928 - Fix building on old distributions, on which man doesn't have a
930 - Fix RAPI not to ignore the MAC address on instance creation
931 - Implement the old instance creation format in the RAPI client.
937 *(Released Thu, 01 Jul 2010)*
939 A small bugfix release:
940 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
941 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
942 a non-existing option
943 - Fix instance create: the --force-variant option was ignored
944 - Improve pylint 0.21 compatibility and warnings with Python 2.6
945 - Fix modify node storage with non-FQDN arguments
946 - Fix RAPI client to authenticate under Python 2.6 when used
947 for more than 5 requests needing authentication
948 - Fix gnt-instance modify -t (storage) giving a wrong error message
949 when converting a non-shutdown drbd instance to plain
955 *(Released Fri, 18 Jun 2010)*
957 A small bugfix release:
959 - Fix live migration of KVM instances started with older Ganeti
960 versions which had fewer hypervisor parameters
961 - Fix gnt-instance grow-disk on down instances
962 - Fix an error-reporting bug during instance migration
963 - Better checking of the ``--net`` and ``--disk`` values, to avoid
964 silently ignoring broken ones
965 - Fix an RPC error reporting bug affecting, for example, RAPI client
967 - Fix bug triggered by different API version os-es on different nodes
968 - Fix a bug in instance startup with custom hvparams: OS level
969 parameters would fail to be applied.
970 - Fix the RAPI client under Python 2.6 (but more work is needed to
971 make it work completely well with OpenSSL)
972 - Fix handling of errors when resolving names from DNS
978 *(Released Thu, 3 Jun 2010)*
980 A medium sized development cycle. Some new features, and some
981 fixes/small improvements/cleanups.
986 The node deamon now tries to mlock itself into memory, unless the
987 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
988 its logs, and falls back to console logging. This allows emergency
989 features such as ``gnt-node powercycle`` to work even in the event of a
990 broken node disk (tested offlining the disk hosting the node's
991 filesystem and dropping its memory caches; don't try this at home)
993 KVM: add vhost-net acceleration support. It can be tested with a new
994 enough version of the kernel and of qemu-kvm.
996 KVM: Add instance chrooting feature. If you use privilege dropping for
997 your VMs you can also now force them to chroot to an empty directory,
998 before starting the emulated guest.
1000 KVM: Add maximum migration bandwith and maximum downtime tweaking
1001 support (requires a new-enough version of qemu-kvm).
1003 Cluster verify will now warn if the master node doesn't have the master
1004 ip configured on it.
1006 Add a new (incompatible) instance creation request format to RAPI which
1007 supports all parameters (previously only a subset was supported, and it
1008 wasn't possible to extend the old format to accomodate all the new
1009 features. The old format is still supported, and a client can check for
1010 this feature, before using it, by checking for its presence in the
1011 ``features`` RAPI resource.
1013 Now with ancient latin support. Try it passing the ``--roman`` option to
1014 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1015 (requires the python-roman module to be installed, in order to work).
1020 As usual many internal code refactorings, documentation updates, and
1023 - Lots of improvements and cleanups to the experimental Remote API
1024 (RAPI) client library.
1025 - A new unit test suite for the core daemon libraries.
1026 - A fix to creating missing directories makes sure the umask is not
1027 applied anymore. This enforces the same directory permissions
1029 - Better handling terminating daemons with ctrl+c (used when running
1030 them in debugging mode).
1031 - Fix a race condition in live migrating a KVM instance, when stat()
1032 on the old proc status file returned EINVAL, which is an unexpected
1034 - Fixed manpage checking with newer man and utf-8 charachters. But now
1035 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1041 *(Released Fri, 7 May 2010)*
1043 Fix a bug which prevented untagged KVM instances from starting.
1049 *(Released Fri, 7 May 2010)*
1051 Another release with a long development cycle, during which many
1052 different features were added.
1054 Significant features
1055 ~~~~~~~~~~~~~~~~~~~~
1057 The KVM hypervisor now can run the individual instances as non-root, to
1058 reduce the impact of a VM being hijacked due to bugs in the
1059 hypervisor. It is possible to run all instances as a single (non-root)
1060 user, to manually specify a user for each instance, or to dynamically
1061 allocate a user out of a cluster-wide pool to each instance, with the
1062 guarantee that no two instances will run under the same user ID on any
1065 An experimental RAPI client library, that can be used standalone
1066 (without the other Ganeti libraries), is provided in the source tree as
1067 ``lib/rapi/client.py``. Note this client might change its interface in
1068 the future, as we iterate on its capabilities.
1070 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1071 replace the cluster's certificates and crypto keys. This might help in
1072 case they have been compromised, or have simply expired.
1074 A new disk option for instance creation has been added that allows one
1075 to "adopt" currently existing logical volumes, with data
1076 preservation. This should allow easier migration to Ganeti from
1077 unmanaged (or managed via other software) instances.
1079 Another disk improvement is the possibility to convert between redundant
1080 (DRBD) and plain (LVM) disk configuration for an instance. This should
1081 allow better scalability (starting with one node and growing the
1082 cluster, or shrinking a two-node cluster to one node).
1084 A new feature that could help with automated node failovers has been
1085 implemented: if a node sees itself as offline (by querying the master
1086 candidates), it will try to shutdown (hard) all instances and any active
1087 DRBD devices. This reduces the risk of duplicate instances if an
1088 external script automatically failovers the instances on such nodes. To
1089 enable this, the cluster parameter ``maintain_node_health`` should be
1090 enabled; in the future this option (per the name) will enable other
1091 automatic maintenance features.
1093 Instance export/import now will reuse the original instance
1094 specifications for all parameters; that means exporting an instance,
1095 deleting it and the importing it back should give an almost identical
1096 instance. Note that the default import behaviour has changed from
1097 before, where it created only one NIC; now it recreates the original
1100 Cluster verify has added a few new checks: SSL certificates validity,
1101 /etc/hosts consistency across the cluster, etc.
1106 As usual, many internal changes were done, documentation fixes,
1109 - Fixed cluster initialization with disabled cluster storage (regression
1110 introduced in 2.1.1)
1111 - File-based storage supports growing the disks
1112 - Fixed behaviour of node role changes
1113 - Fixed cluster verify for some corner cases, plus a general rewrite of
1114 cluster verify to allow future extension with more checks
1115 - Fixed log spamming by watcher and node daemon (regression introduced
1117 - Fixed possible validation issues when changing the list of enabled
1119 - Fixed cleanup of /etc/hosts during node removal
1120 - Fixed RAPI response for invalid methods
1121 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1122 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1123 from ide disks, etc.)
1124 - Allow OS changes without re-installation (to record a changed OS
1125 outside of Ganeti, or to allow OS renames)
1126 - Allow instance creation without OS installation (useful for example if
1127 the OS will be installed manually, or restored from a backup not in
1129 - Implemented option to make cluster ``copyfile`` use the replication
1131 - Added list of enabled hypervisors to ssconf (possibly useful for
1133 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1135 - A partial form of node re-IP is possible via node readd, which now
1136 allows changed node primary IP
1137 - Command line utilities now show an informational message if the job is
1139 - The logs of the master daemon now show the PID/UID/GID of the
1146 *(Released Fri, 12 Mar 2010)*
1148 During the 2.1.0 long release candidate cycle, a lot of improvements and
1149 changes have accumulated with were released later as 2.1.1.
1154 The node evacuate command (``gnt-node evacuate``) was significantly
1155 rewritten, and as such the IAllocator protocol was changed - a new
1156 request type has been added. This unfortunate change during a stable
1157 series is designed to improve performance of node evacuations; on
1158 clusters with more than about five nodes and which are well-balanced,
1159 evacuation should proceed in parallel for all instances of the node
1160 being evacuated. As such, any existing IAllocator scripts need to be
1161 updated, otherwise the above command will fail due to the unknown
1162 request. The provided "dumb" allocator has not been updated; but the
1163 ganeti-htools package supports the new protocol since version 0.2.4.
1165 Another important change is increased validation of node and instance
1166 names. This might create problems in special cases, if invalid host
1167 names are being used.
1169 Also, a new layer of hypervisor parameters has been added, that sits at
1170 OS level between the cluster defaults and the instance ones. This allows
1171 customisation of virtualization parameters depending on the installed
1172 OS. For example instances with OS 'X' may have a different KVM kernel
1173 (or any other parameter) than the cluster defaults. This is intended to
1174 help managing a multiple OSes on the same cluster, without manual
1175 modification of each instance's parameters.
1177 A tool for merging clusters, ``cluster-merge``, has been added in the
1178 tools sub-directory.
1183 - Improved the int/float conversions that should make the code more
1184 robust in face of errors from the node daemons
1185 - Fixed the remove node code in case of internal configuration errors
1186 - Fixed the node daemon behaviour in face of inconsistent queue
1187 directory (e.g. read-only file-system where we can't open the files
1189 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1190 now it either aborts cleanly or, if given the new "auto_promote"
1191 parameter, will automatically promote other nodes as needed
1192 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1193 completely prevent Ganeti from working
1194 - Fixed bug for instance export when not all disks were successfully
1196 - Fixed behaviour of node add when the new node is slow in starting up
1198 - Fixed handling of signals in the LUXI client, which should improve
1199 behaviour of command-line scripts
1200 - Added checks for invalid node/instance names in the configuration (now
1201 flagged during cluster verify)
1202 - Fixed watcher behaviour for disk activation errors
1203 - Fixed two potentially endless loops in http library, which led to the
1204 RAPI daemon hanging and consuming 100% CPU in some cases
1205 - Fixed bug in RAPI daemon related to hashed passwords
1206 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1208 - Enhanced compatibility with non-Debian OSes, but not using absolute
1209 path in some commands and allowing customisation of the ssh
1210 configuration directory
1211 - Fixed possible future issue with new Python versions by abiding to the
1212 proper use of ``__slots__`` attribute on classes
1213 - Added checks that should prevent directory traversal attacks
1214 - Many documentation fixes based on feedback from users
1219 - Added an "early_release" more for instance replace disks and node
1220 evacuate, where we release locks earlier and thus allow higher
1221 parallelism within the cluster
1222 - Added watcher hooks, intended to allow the watcher to restart other
1223 daemons (e.g. from the ganeti-nbma project), but they can be used of
1224 course for any other purpose
1225 - Added a compile-time disable for DRBD barriers, to increase
1226 performance if the administrator trusts the power supply or the
1227 storage system to not lose writes
1228 - Added the option of using syslog for logging instead of, or in
1229 addition to, Ganeti's own log files
1230 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1231 can indeed boot from a paravirtual NIC
1232 - Added a generic debug level for many operations; while this is not
1233 used widely yet, it allows one to pass the debug value all the way to
1235 - Enhanced the hooks environment for instance moves (failovers,
1236 migrations) where the primary/secondary nodes changed during the
1237 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1238 - Enhanced data validations for many user-supplied values; one important
1239 item is the restrictions imposed on instance and node names, which
1240 might reject some (invalid) host names
1241 - Add a configure-time option to disable file-based storage, if it's not
1242 needed; this allows greater security separation between the master
1243 node and the other nodes from the point of view of the inter-node RPC
1245 - Added user notification in interactive tools if job is waiting in the
1246 job queue or trying to acquire locks
1247 - Added log messages when a job is waiting for locks
1248 - Added filtering by node tags in instance operations which admit
1249 multiple instances (start, stop, reboot, reinstall)
1250 - Added a new tool for cluster mergers, ``cluster-merge``
1251 - Parameters from command line which are of the form ``a=b,c=d`` can now
1252 use backslash escapes to pass in values which contain commas,
1253 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1255 - For KVM, the instance name is the first parameter passed to KVM, so
1256 that it's more visible in the process list
1262 *(Released Tue, 2 Mar 2010)*
1264 Ganeti 2.1 brings many improvements with it. Major changes:
1266 - Added infrastructure to ease automated disk repairs
1267 - Added new daemon to export configuration data in a cheaper way than
1268 using the remote API
1269 - Instance NICs can now be routed instead of being associated with a
1271 - Improved job locking logic to reduce impact of jobs acquiring multiple
1272 locks waiting for other long-running jobs
1274 In-depth implementation details can be found in the Ganeti 2.1 design
1280 - Added chroot hypervisor
1281 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1283 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1284 ``bootloader_path`` and ``bootloader_args``)
1285 - Added the ``use_localtime`` option for the xen-hvm and kvm
1286 hypervisors, and the default value for this has changed to false (in
1287 2.0 xen-hvm always enabled it)
1288 - Added luxi call to submit multiple jobs in one go
1289 - Added cluster initialization option to not modify ``/etc/hosts``
1291 - Added network interface parameters
1292 - Added dry run mode to some LUs
1293 - Added RAPI resources:
1295 - ``/2/instances/[instance_name]/info``
1296 - ``/2/instances/[instance_name]/replace-disks``
1297 - ``/2/nodes/[node_name]/evacuate``
1298 - ``/2/nodes/[node_name]/migrate``
1299 - ``/2/nodes/[node_name]/role``
1300 - ``/2/nodes/[node_name]/storage``
1301 - ``/2/nodes/[node_name]/storage/modify``
1302 - ``/2/nodes/[node_name]/storage/repair``
1304 - Added OpCodes to evacuate or migrate all instances on a node
1305 - Added new command to list storage elements on nodes (``gnt-node
1306 list-storage``) and modify them (``gnt-node modify-storage``)
1307 - Added new ssconf files with master candidate IP address
1308 (``ssconf_master_candidates_ips``), node primary IP address
1309 (``ssconf_node_primary_ips``) and node secondary IP address
1310 (``ssconf_node_secondary_ips``)
1311 - Added ``ganeti-confd`` and a client library to query the Ganeti
1312 configuration via UDP
1313 - Added ability to run hooks after cluster initialization and before
1315 - Added automatic mode for disk replace (``gnt-instance replace-disks
1317 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1318 after catastrophic data-loss
1319 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1321 - Added ``gnt-instance move`` command to move instances
1322 - Added ``gnt-cluster watcher`` command to control watcher
1323 - Added ``gnt-node powercycle`` command to powercycle nodes
1324 - Added new job status field ``lock_status``
1325 - Added parseable error codes to cluster verification (``gnt-cluster
1326 verify --error-codes``) and made output less verbose (use
1327 ``--verbose`` to restore previous behaviour)
1328 - Added UUIDs to the main config entities (cluster, nodes, instances)
1329 - Added support for OS variants
1330 - Added support for hashed passwords in the Ganeti remote API users file
1332 - Added option to specify maximum timeout on instance shutdown
1333 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1334 - Added new helper script to start and stop Ganeti daemons
1335 (``daemon-util``), with the intent to reduce the work necessary to
1336 adjust Ganeti for non-Debian distributions and to start/stop daemons
1338 - Added more unittests
1339 - Fixed critical bug in ganeti-masterd startup
1340 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1341 now customisable at the cluster level for both the KVM and Xen
1342 hypervisors using the new ``migration_port`` parameter
1343 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1344 reinstalling an instance
1345 - Allowed ``@`` in tag names
1346 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1347 - Many documentation updates
1348 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1349 - ``gnt-instance reinstall`` can now reinstall multiple instances
1350 - Updated many command line parameters
1351 - Introduced new OS API version 15
1352 - No longer support a default hypervisor
1353 - Treat virtual LVs as inexistent
1354 - Improved job locking logic to reduce lock contention
1355 - Match instance and node names case insensitively
1356 - Reimplemented bash completion script to be more complete
1363 *(Released Thu, 4 Feb 2010)*
1365 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1366 - Fix a string formatting bug
1367 - Improve safety of the code in some error paths
1368 - Improve data validation in the master of values returned from nodes
1374 *(Released Thu, 17 Dec 2009)*
1376 - Fix security issue due to missing validation of iallocator names; this
1377 allows local and remote execution of arbitrary executables
1378 - Fix failure of gnt-node list during instance removal
1379 - Ship the RAPI documentation in the archive
1385 *(Released Wed, 30 Sep 2009)*
1387 - Fixed many wrong messages
1388 - Fixed a few bugs related to the locking library
1389 - Fixed MAC checking at instance creation time
1390 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1391 - Fixed a few issues related to signal handling in both daemons and
1393 - Fixed the example startup script provided
1394 - Fixed insserv dependencies in the example startup script (patch from
1396 - Fixed handling of drained nodes in the iallocator framework
1397 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1399 - Fixed error related to invalid job IDs in job polling
1400 - Fixed job/opcode persistence on unclean master shutdown
1401 - Fixed handling of partial job processing after unclean master
1403 - Fixed error reporting from LUs, previously all errors were converted
1404 into execution errors
1405 - Fixed error reporting from burnin
1406 - Decreased significantly the memory usage of the job queue
1407 - Optimised slightly multi-job submission
1408 - Optimised slightly opcode loading
1409 - Backported the multi-job submit framework from the development
1410 branch; multi-instance start and stop should be faster
1411 - Added script to clean archived jobs after 21 days; this will reduce
1412 the size of the queue directory
1413 - Added some extra checks in disk size tracking
1414 - Added an example ethers hook script
1415 - Added a cluster parameter that prevents Ganeti from modifying of
1417 - Added more node information to RAPI responses
1418 - Added a ``gnt-job watch`` command that allows following the ouput of a
1420 - Added a bind-address option to ganeti-rapi
1421 - Added more checks to the configuration verify
1422 - Enhanced the burnin script such that some operations can be retried
1424 - Converted instance reinstall to multi-instance model
1430 *(Released Fri, 7 Aug 2009)*
1432 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1433 to allow using the pre-2.0.2 behaviour in activation, if any existing
1434 instances have mismatched disk sizes in the configuration
1435 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1436 any configuration mismatches for disk sizes
1437 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1438 failover to work on two-node clusters
1439 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1441 - Fixed detection of OS script errors in ``gnt-backup export``
1442 - Fixed exit code of ``gnt-backup export``
1448 *(Released Fri, 17 Jul 2009)*
1450 - Added experimental support for stripped logical volumes; this should
1451 enhance performance but comes with a higher complexity in the block
1452 device handling; stripping is only enabled when passing
1453 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1454 affected even in the non-stripped mode
1455 - Improved resiliency against transient failures at the end of DRBD
1456 resyncs, and in general of DRBD resync checks
1457 - Fixed a couple of issues with exports and snapshot errors
1458 - Fixed a couple of issues in instance listing
1459 - Added display of the disk size in ``gnt-instance info``
1460 - Fixed checking for valid OSes in instance creation
1461 - Fixed handling of the "vcpus" parameter in instance listing and in
1462 general of invalid parameters
1463 - Fixed http server library, and thus RAPI, to handle invalid
1464 username/password combinations correctly; this means that now they
1465 report unauthorized for queries too, not only for modifications,
1466 allowing earlier detect of configuration problems
1467 - Added a new "role" node list field, equivalent to the master/master
1468 candidate/drained/offline flags combinations
1469 - Fixed cluster modify and changes of candidate pool size
1470 - Fixed cluster verify error messages for wrong files on regular nodes
1471 - Fixed a couple of issues with node demotion from master candidate role
1472 - Fixed node readd issues
1473 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1474 - Added a new ``--no-voting`` option for masterfailover to fix failover
1475 on two-nodes clusters when the former master node is unreachable
1476 - Added instance reinstall over RAPI
1482 *(Released Tue, 16 Jun 2009)*
1484 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1485 allow re-adding the start in single-user option (regression from 1.2)
1486 - the watcher writes the instance status to a file, to allow monitoring
1487 to report the instance status (from the master) based on cached
1488 results of the watcher's queries; while this can get stale if the
1489 watcher is being locked due to other work on the cluster, this is
1490 still an improvement
1491 - the watcher now also restarts the node daemon and the rapi daemon if
1493 - fixed the watcher to handle full and drained queue cases
1494 - hooks export more instance data in the environment, which helps if
1495 hook scripts need to take action based on the instance's properties
1496 (no longer need to query back into ganeti)
1497 - instance failovers when the instance is stopped do not check for free
1498 RAM, so that failing over a stopped instance is possible in low memory
1500 - rapi uses queries for tags instead of jobs (for less job traffic), and
1501 for cluster tags it won't talk to masterd at all but read them from
1503 - a couple of error handling fixes in RAPI
1504 - drbd handling: improved the error handling of inconsistent disks after
1505 resync to reduce the frequency of "there are some degraded disks for
1506 this instance" messages
1507 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1508 error handling path called a wrong function name)
1514 *(Released Wed, 27 May 2009)*
1516 - no changes from rc5
1522 *(Released Wed, 20 May 2009)*
1524 - fix a couple of bugs (validation, argument checks)
1525 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1526 - some small improvements to RAPI and IAllocator
1527 - make watcher automatically start the master daemon if down
1533 *(Released Mon, 27 Apr 2009)*
1535 - change the OS list to not require locks; this helps with big clusters
1536 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1537 volume group is broken
1538 - ``gnt-instance info``, without any arguments, doesn't run for all
1539 instances anymore; either pass ``--all`` or pass the desired
1540 instances; this helps against mistakes on big clusters where listing
1541 the information for all instances takes a long time
1542 - miscellaneous doc and man pages fixes
1548 *(Released Wed, 8 Apr 2009)*
1550 - Change the internal locking model of some ``gnt-node`` commands, in
1551 order to reduce contention (and blocking of master daemon) when
1552 batching many creation/reinstall jobs
1553 - Fixes to Xen soft reboot
1554 - No longer build documentation at build time, instead distribute it in
1555 the archive, in order to reduce the need for the whole docbook/rst
1562 *(Released Fri, 27 Mar 2009)*
1564 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1565 - Fix watcher startup sequence, improves the behaviour of busy clusters
1566 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1567 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1568 bind address and other places
1569 - Some documentation fixes and updates
1575 *(Released Mon, 2 Mar 2009)*
1577 - More documentation updates, now all docs should be more-or-less
1579 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1581 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1583 - Fix an issue related to $libdir/run/ganeti and cluster creation
1589 *(Released Thu, 19 Feb 2009)*
1591 - Xen PVM and KVM have switched the default value for the instance root
1592 disk to the first partition on the first drive, instead of the whole
1593 drive; this means that the OS installation scripts must be changed
1595 - Man pages have been updated
1596 - RAPI has been switched by default to HTTPS, and the exported functions
1597 should all work correctly
1598 - RAPI v1 has been removed
1599 - Many improvements to the KVM hypervisor
1600 - Block device errors are now better reported
1601 - Many other bugfixes and small improvements
1607 *(Released Mon, 26 Jan 2009)*
1609 - Version 2 is a general rewrite of the code and therefore the
1610 differences are too many to list, see the design document for 2.0 in
1611 the ``doc/`` subdirectory for more details
1612 - In this beta version there is not yet a migration path from 1.2 (there
1613 will be one in the final 2.0 release)
1614 - A few significant changes are:
1616 - all commands are executed by a daemon (``ganeti-masterd``) and the
1617 various ``gnt-*`` commands are just front-ends to it
1618 - all the commands are entered into, and executed from a job queue,
1619 see the ``gnt-job(8)`` manpage
1620 - the RAPI daemon supports read-write operations, secured by basic
1621 HTTP authentication on top of HTTPS
1622 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1623 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1624 to migrate to DRBD 8 first while still running Ganeti 1.2)
1625 - DRBD devices are using statically allocated minor numbers, which
1626 will be assigned to existing instances during the migration process
1627 - there is support for both Xen PVM and Xen HVM instances running on
1629 - KVM virtualization is supported too
1630 - file-based storage has been implemented, which means that it is
1631 possible to run the cluster without LVM and DRBD storage, for
1632 example using a shared filesystem exported from shared storage (and
1633 still have live migration)
1639 *(Released Tue, 13 Jan 2009)*
1641 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1642 - Reuse the old instance mac address by default on instance import, if
1643 the instance name is the same.
1644 - Handle situations in which the node info rpc returns incomplete
1646 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1647 - Improved version of batcher:
1649 - state file support
1650 - instance mac address support
1651 - support for HVM clusters/instances
1653 - Add an option to show the number of cpu sockets and nodes in
1655 - Support OSes that handle more than one version of the OS api (but do
1656 not change the current API in any other way)
1657 - Fix ``gnt-node migrate``
1658 - ``gnt-debug`` man page
1659 - Fixes various more typos and small issues
1660 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1666 *(Released Wed, 24 Sep 2008)*
1668 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1669 type of disk exported to fully virtualized instances.
1670 - provide access to the serial console of HVM instances
1671 - instance auto_balance flag, set by default. If turned off it will
1672 avoid warnings on cluster verify if there is not enough memory to fail
1673 over an instance. in the future it will prevent automatically failing
1674 it over when we will support that.
1675 - batcher tool for instance creation, see ``tools/README.batcher``
1676 - ``gnt-instance reinstall --select-os`` to interactively select a new
1677 operating system when reinstalling an instance.
1678 - when changing the memory amount on instance modify a check has been
1679 added that the instance will be able to start. also warnings are
1680 emitted if the instance will not be able to fail over, if auto_balance
1682 - documentation fixes
1683 - sync fields between ``gnt-instance list/modify/add/import``
1684 - fix a race condition in drbd when the sync speed was set after giving
1685 the device a remote peer.
1691 *(Released Tue, 22 Jul 2008)*
1693 - note: the allowed size and number of tags per object were reduced
1694 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1695 - fixed twisted 8.x compatibility
1696 - fixed ``gnt-instance replace-disks`` with iallocator
1697 - add TCP keepalives on twisted connections to detect restarted nodes
1698 - disk increase support, see ``gnt-instance grow-disk``
1699 - implement bulk node/instance query for RAPI
1700 - add tags in node/instance listing (optional)
1701 - experimental migration (and live migration) support, read the man page
1702 for ``gnt-instance migrate``
1703 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1704 has some small improvements in handling its state file
1710 *(Released Fri, 13 Jun 2008)*
1712 - Experimental readonly, REST-based remote API implementation;
1713 automatically started on master node, TCP port 5080, if enabled by
1714 ``--enable-rapi`` parameter to configure script.
1715 - Instance allocator support. Add and import instance accept a
1716 ``--iallocator`` parameter, and call that instance allocator to decide
1717 which node to use for the instance. The iallocator document describes
1718 what's expected from an allocator script.
1719 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1720 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1721 node is lost there is still enough memory to fail over the instances
1723 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1724 ``gnt-cluster verify``, to check for site-specific compliance. All the
1725 hooks will run, and their output, if any, will be displayed. Any
1726 failing hook will make the verification return an error value.
1727 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1728 primary and secondary interfaces
1729 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1730 that is still declared as part of the cluster and has failed.
1731 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1732 specifying output fields, that just adds the chosen fields to the
1734 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1735 export from the filesystem.
1736 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1737 have been added. Using them you can enable/disable acpi and pae
1738 support, and specify a path for a cd image to be exported to the
1739 instance. These parameters as the name suggest only work on HVM
1741 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1742 PAE support will be set to the previously hardcoded values, but the
1743 (previously hardcoded) path to the CDROM ISO image will be unset and
1744 if required, needs to be set manually with ``gnt-instance modify``
1746 - The address to which an instance's VNC console is bound is now
1747 selectable per-instance, rather than being cluster wide. Of course
1748 this only applies to instances controlled via VNC, so currently just
1749 applies to HVM clusters.
1755 *(Released Mon, 18 Feb 2008)*
1757 - more tweaks to the disk activation code (especially helpful for DRBD)
1758 - change the default ``gnt-instance list`` output format, now there is
1759 one combined status field (see the manpage for the exact values this
1761 - some more fixes for the mac export to hooks change
1762 - make Ganeti not break with DRBD 8.2.x (which changed the version
1763 format in ``/proc/drbd``) (issue 24)
1764 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1765 template, allowing migration from DRBD0.7+MD to DRBD8
1771 *(Released Wed, 30 Jan 2008)*
1773 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1775 - add command aliases infrastructure and a few aliases
1776 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1777 man pages and the ``--help`` option of ``gnt-node
1778 list``/``gnt-instance list``
1779 - fix ``gnt-backup list`` with down nodes (issue 21)
1780 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1781 - fix the dist archive and add a check for including svn/git files in
1783 - some developer-related changes: improve the burnin and the QA suite,
1784 add an upload script for testing during development
1790 *(Released Wed, 16 Jan 2008)*
1792 - experimental HVM support, read the install document, section
1793 "Initializing the cluster"
1794 - allow for the PVM hypervisor per-instance kernel and initrd paths
1795 - add a new command ``gnt-cluster verify-disks`` which uses a new
1796 algorithm to improve the reconnection of the DRBD pairs if the device
1797 on the secondary node has gone away
1798 - make logical volume code auto-activate LVs at disk activation time
1799 - slightly improve the speed of activating disks
1800 - allow specification of the MAC address at instance creation time, and
1801 changing it later via ``gnt-instance modify``
1802 - fix handling of external commands that generate lots of output on
1804 - update documentation with regard to minimum version of DRBD8 supported
1810 *(Released Tue, 4 Dec 2007)*
1812 - Log the ``xm create`` output to the node daemon log on failure (to
1813 help diagnosing the error)
1814 - In debug mode, log all external commands output if failed to the logs
1815 - Change parsing of lvm commands to ignore stderr
1821 *(Released Wed, 28 Nov 2007)*
1823 - Another round of updates to the DRBD 8 code to deal with more failures
1824 in the replace secondary node operation
1825 - Some more logging of failures in disk operations (lvm, drbd)
1826 - A few documentation updates
1833 *(Released Tue, 13 Nov 2007)*
1835 - Change configuration file format from Python's Pickle to JSON.
1836 Upgrading is possible using the cfgupgrade utility.
1837 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1838 faster replace disks and is more stable (DRBD 8 has many improvements
1839 compared to DRBD 0.7)
1840 - Added command line tags support (see man pages for ``gnt-instance``,
1841 ``gnt-node``, ``gnt-cluster``)
1842 - Added instance rename support
1843 - Added multi-instance startup/shutdown
1844 - Added cluster rename support
1845 - Added ``gnt-node evacuate`` to simplify some node operations
1846 - Added instance reboot operation that can speedup reboot as compared to
1848 - Soften the requirement that hostnames are in FQDN format
1849 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1851 - Removed dependency on debian's patched fping that uses the
1852 non-standard ``-S`` option
1853 - Now the OS definitions are searched for in multiple, configurable
1854 paths (easier for distros to package)
1855 - Some changes to the hooks infrastructure (especially the new
1856 post-configuration update hook)
1857 - Other small bugfixes
1859 .. vim: set textwidth=72 syntax=rst :
projects / ganeti-local / blob