8 *(Released Fri, 11 May 2012)*
10 A small bugfix release.
12 The main issues solved are on the topic of compatibility with newer LVM
15 - fixed parsing of ``lv_attr`` field
16 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
17 the ``--force`` flag is needed
19 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
20 to accept kernel 3.x too (was hardcoded to 2.6.*).
22 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
23 of hook results and that also made display of other errors suboptimal
24 was fixed; the code behaves now like 2.4 and earlier.
26 Another change in 2.5, the cleanup of the OS scripts environment, is too
27 aggressive: it removed even the ``PATH`` variable, which requires the OS
28 scripts to *always* need to export it. Since this is a bit too strict,
29 we now export a minimal PATH, the same that we export for hooks.
31 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
32 integrated into this release.
34 Finally, a few other miscellaneous changes were done (no new features,
35 just small improvements):
37 - Fix ``gnt-group --help`` display
38 - Fix hardcoded Xen kernel path
39 - Fix grow-disk handling of invalid units
40 - Update synopsis for ``gnt-cluster repair-disk-sizes``
41 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
47 *(Released Thu, 12 Apr 2012)*
49 Incompatible/important changes and bugfixes
50 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
52 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
53 resource's ``ip_check`` parameter changed from ``True`` to ``False``
54 to match the underlying LUXI interface.
55 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
56 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
57 not maintain backwards-compatibility as the underlying operation
58 changed in an incompatible way. The RAPI client can talk to old
59 servers, but it needs to be told so as the return value changed.
60 - When creating file-based instances via RAPI, the ``file_driver``
61 parameter no longer defaults to ``loop`` and must be specified.
62 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
64 - Support for the undocumented and deprecated RAPI instance creation
65 request format version 0 has been dropped. Use version 1, supported
66 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
67 - Pyparsing 1.4.6 or above is required, see :doc:`installation
68 documentation <install>`.
69 - The "cluster-verify" hooks are now executed per group by the
70 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
71 if you just run ``gnt-cluster verify``, which generates one opcode per
73 - The environment as passed to the OS scripts is cleared, and thus no
74 environment variables defined in the node daemon's environment will be
75 inherited by the scripts.
76 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
78 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
79 support operations involving multiple node groups.
80 - Offline nodes are ignored when failing over an instance.
81 - Support for KVM version 1.0, which changed the version reporting format
83 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
85 - ``Makefile`` is now compatible with Automake 1.11.2
86 - Includes all bugfixes made in the 2.4 series
91 - The ganeti-htools project has been merged into the ganeti-core source
92 tree and will be built as part of Ganeti (see :doc:`install-quick`).
93 - Implemented support for :doc:`shared storage <design-shared-storage>`.
94 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
95 GPT-style partition tables (requires `parted
96 <http://www.gnu.org/s/parted/>`_).
97 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
98 - Allowed adding tags on instance creation.
99 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
100 - Allow instances to be started in a paused state, enabling the user to
101 see the complete console output on boot using the console.
102 - Added new hypervisor flag to control default reboot behaviour
103 (``reboot_behavior``).
104 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
105 - Improved out-of-band management support:
107 - Added ``gnt-node health`` command reporting the health status of
109 - Added ``gnt-node power`` command to manage power status of nodes.
110 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
112 - Instance migration can fall back to failover if instance is not
114 - Filters can be used when listing nodes, instances, groups and locks;
115 see *ganeti(7)* manpage.
116 - Added post-execution status as variables to :doc:`hooks <hooks>`
118 - Instance tags are exported/imported together with the instance.
119 - When given an explicit job ID, ``gnt-job info`` will work for archived
121 - Jobs can define dependencies on other jobs (not yet supported via
122 RAPI or command line, but used by internal commands and usable via
125 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
128 - Instance failover is now available as a RAPI resource
129 (``/2/instances/[instance_name]/failover``).
130 - ``gnt-instance info`` defaults to static information if primary node
132 - Opcodes have a new ``comment`` attribute.
133 - Added basic SPICE support to KVM hypervisor.
134 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
136 Node group improvements
137 ~~~~~~~~~~~~~~~~~~~~~~~
139 - ``gnt-cluster verify`` has been modified to check groups separately,
140 thereby improving performance.
141 - Node group support has been added to ``gnt-cluster verify-disks``,
142 which now operates per node group.
143 - Watcher has been changed to work better with node groups.
145 - One process and state file per node group.
146 - Slow watcher in one group doesn't block other group's watcher.
148 - Added new command, ``gnt-group evacuate``, to move all instances in a
149 node group to other groups.
150 - Added ``gnt-instance change-group`` to move an instance to another
152 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
153 per-group operations.
154 - Node groups can be tagged.
155 - Some operations switch from an exclusive to a shared lock as soon as
157 - Instance's primary and secondary nodes' groups are now available as
158 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
159 and ``snodes.group.uuid``).
164 - Numerous updates to documentation and manpages.
166 - :doc:`RAPI <rapi>` documentation now has detailed parameter
168 - Some opcode/job results are now also documented, see :doc:`RAPI
171 - A lockset's internal lock is now also visible in lock monitor.
172 - Log messages from job queue workers now contain information about the
173 opcode they're processing.
174 - ``gnt-instance console`` no longer requires the instance lock.
175 - A short delay when waiting for job changes reduces the number of LUXI
176 requests significantly.
177 - DRBD metadata volumes are overwritten with zeros during disk creation.
178 - Out-of-band commands no longer acquire the cluster lock in exclusive
180 - ``devel/upload`` now uses correct permissions for directories.
186 *(Released Fri, 23 Mar 2012)*
188 This was the sixth release candidate of the 2.5 series.
194 *(Released Mon, 9 Jan 2012)*
196 This was the fifth release candidate of the 2.5 series.
202 *(Released Thu, 27 Oct 2011)*
204 This was the fourth release candidate of the 2.5 series.
210 *(Released Wed, 26 Oct 2011)*
212 This was the third release candidate of the 2.5 series.
218 *(Released Tue, 18 Oct 2011)*
220 This was the second release candidate of the 2.5 series.
226 *(Released Tue, 4 Oct 2011)*
228 This was the first release candidate of the 2.5 series.
234 *(Released Wed, 31 Aug 2011)*
236 This was the third beta release of the 2.5 series.
242 *(Released Mon, 22 Aug 2011)*
244 This was the second beta release of the 2.5 series.
250 *(Released Fri, 12 Aug 2011)*
252 This was the first beta release of the 2.5 series.
258 *(Released Thu, 27 Oct 2011)*
260 - Fixed bug when parsing command line parameter values ending in
262 - Fixed assertion error after unclean master shutdown
263 - Disable HTTP client pool for RPC, significantly reducing memory usage
265 - Fixed queue archive creation with wrong permissions
271 *(Released Tue, 23 Aug 2011)*
275 - Fixed documentation for importing with ``--src-dir`` option
276 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
277 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
283 *(Released Fri, 5 Aug 2011)*
285 Many bug-fixes and a few small features:
287 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
288 issues when you tried to add an instance with two MAC addresses in one
290 - KVM: fixed per-instance stored UID value
291 - KVM: configure bridged NICs at migration start
292 - KVM: Fix a bug where instance will not start with never KVM versions
294 - Added OS search path to ``gnt-cluster info``
295 - Fixed an issue with ``file_storage_dir`` where you were forced to
296 provide an absolute path, but the documentation states it is a
297 relative path, the documentation was right
298 - Added a new parameter to instance stop/start called ``--no-remember``
299 that will make the state change to not be remembered
300 - Implemented ``no_remember`` at RAPI level
301 - Improved the documentation
302 - Node evacuation: don't call IAllocator if node is already empty
303 - Fixed bug in DRBD8 replace disks on current nodes
304 - Fixed bug in recreate-disks for DRBD instances
305 - Moved assertion checking locks in ``gnt-instance replace-disks``
306 causing it to abort with not owning the right locks for some situation
307 - Job queue: Fixed potential race condition when cancelling queued jobs
308 - Fixed off-by-one bug in job serial generation
309 - ``gnt-node volumes``: Fix instance names
310 - Fixed aliases in bash completion
311 - Fixed a bug in reopening log files after being sent a SIGHUP
312 - Added a flag to burnin to allow specifying VCPU count
313 - Bugfixes to non-root Ganeti configuration
319 *(Released Thu, 12 May 2011)*
321 Many bug-fixes and a few new small features:
323 - Fixed a bug related to log opening failures
324 - Fixed a bug in instance listing with orphan instances
325 - Fixed a bug which prevented resetting the cluster-level node parameter
326 ``oob_program`` to the default
327 - Many fixes related to the ``cluster-merge`` tool
328 - Fixed a race condition in the lock monitor, which caused failures
329 during (at least) creation of many instances in parallel
330 - Improved output for gnt-job info
331 - Removed the quiet flag on some ssh calls which prevented debugging
333 - Improved the N+1 failure messages in cluster verify by actually
334 showing the memory values (needed and available)
335 - Increased lock attempt timeouts so that when executing long operations
336 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
337 too early and thus prevent the use of the 'fair' mechanism
338 - Changed instance query data (``gnt-instance info``) to not acquire
339 locks unless needed, thus allowing its use on locked instance if only
340 static information is asked for
341 - Improved behaviour with filesystems that do not support rename on an
343 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
344 kept locks on all nodes during the wipe, which is unneeded
345 - Fixed ``gnt-watcher`` handling of errors during hooks execution
346 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
347 10GiB) which caused the wipe to fail right at the end in some cases
348 - Fixed master IP activation when doing master failover with no-voting
349 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
350 the master node itself
351 - Fixed potential data-loss in under disk full conditions, where Ganeti
352 wouldn't check correctly the return code and would consider
353 partially-written files 'correct'
354 - Fixed bug related to multiple VGs and DRBD disk replacing
355 - Added new disk parameter ``metavg`` that allows placement of the meta
356 device for DRBD in a different volume group
357 - Fixed error handling in the node daemon when the system libc doesn't
358 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
359 - Fixed lock release during replace-disks, which kept cluster-wide locks
360 when doing disk replaces with an iallocator script
361 - Added check for missing bridges in cluster verify
362 - Handle EPIPE errors while writing to the terminal better, so that
363 piping the output to e.g. ``less`` doesn't cause a backtrace
364 - Fixed rare case where a ^C during Luxi calls could have been
365 interpreted as server errors, instead of simply terminating
366 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
368 - Added a few more parameters to the KVM hypervisor, allowing a second
369 CDROM, custom disk type for CDROMs and a floppy image
370 - Removed redundant message in instance rename when the name is given
372 - Added option to ``gnt-instance recreate-disks`` to allow creating the
373 disks on new nodes, allowing recreation when the original instance
374 nodes are completely gone
375 - Added option when converting disk templates to DRBD to skip waiting
376 for the resync, in order to make the instance available sooner
377 - Added two new variables to the OS scripts environment (containing the
379 - Made the root_path and optional parameter for the xen-pvm hypervisor,
380 to allow use of ``pvgrub`` as bootloader
381 - Changed the instance memory modifications to only check out-of-memory
382 conditions on memory increases, and turned the secondary node warnings
383 into errors (they can still be overridden via ``--force``)
384 - Fixed the handling of a corner case when the Python installation gets
385 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
386 to execute a command that doesn't exist
387 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
388 node of the instance returned failures during instance shutdown; this
389 adds the option ``--ignore-consistency`` to gnt-instance move
391 And as usual, various improvements to the error messages, documentation
397 *(Released Wed, 09 Mar 2011)*
399 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
400 the RAPI users file if run twice (even with ``--dry-run``).
402 The release fixes that bug (nothing else changed).
408 *(Released Mon, 07 Mar 2011)*
410 Final 2.4.0 release. Just a few small fixes:
412 - Fixed RAPI node evacuate
413 - Fixed the kvm-ifup script
414 - Fixed internal error handling for special job cases
415 - Updated man page to specify the escaping feature for options
421 *(Released Mon, 28 Feb 2011)*
423 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
424 that this feature wiped the disks of the wrong instance, leading to loss
429 - Fixed title of query field containing instance name
430 - Expanded the glossary in the documentation
431 - Fixed one unittest (internal issue)
437 *(Released Mon, 21 Feb 2011)*
439 A number of bug fixes plus just a couple functionality changes.
441 On the user-visible side, the ``gnt-* list`` command output has changed
442 with respect to "special" field states. The current rc1 style of display
443 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
444 the default output mode special fields states are displayed as follows:
446 - Offline resource: ``*``
447 - Unavailable/not applicable: ``-``
448 - Data missing (RPC failure): ``?``
449 - Unknown field: ``??``
451 Another user-visible change is the addition of ``--force-join`` to
456 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
457 - Fixed regression in RAPI/instance reinstall where all parameters were
458 required (instead of optional)
459 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
460 - Fixed iallocator usage (offline nodes were not considered offline)
461 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
462 - Fixed hypervisor and OS parameter validation with respect to
464 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
466 - Fixed ``tools/listrunner`` with respect to agent-based usage
472 *(Released Fri, 4 Feb 2011)*
474 Many changes and fixes since the beta1 release. While there were some
475 internal changes, the code has been mostly stabilised for the RC
478 Note: the dumb allocator was removed in this release, as it was not kept
479 up-to-date with the IAllocator protocol changes. It is recommended to
480 use the ``hail`` command from the ganeti-htools package.
482 Note: the 2.4 and up versions of Ganeti are not compatible with the
483 0.2.x branch of ganeti-htools. You need to upgrade to
484 ganeti-htools-0.3.0 (or later).
486 Regressions fixed from 2.3
487 ~~~~~~~~~~~~~~~~~~~~~~~~~~
489 - Fixed the ``gnt-cluster verify-disks`` command
490 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
492 - Fixed disk adoption breakage
493 - Fixed wrong headers in instance listing for field aliases
498 - Fixed corner case in KVM handling of NICs
499 - Fixed many cases of wrong handling of non-vm_capable nodes
500 - Fixed a bug where a missing instance symlink was not possible to
501 recreate with any ``gnt-*`` command (now ``gnt-instance
502 activate-disks`` does it)
503 - Fixed the volume group name as reported by ``gnt-cluster
505 - Increased timeouts for the import-export code, hopefully leading to
506 fewer aborts due network or instance timeouts
507 - Fixed bug in ``gnt-node list-storage``
508 - Fixed bug where not all daemons were started on cluster
509 initialisation, but only at the first watcher run
510 - Fixed many bugs in the OOB implementation
511 - Fixed watcher behaviour in presence of instances with offline
513 - Fixed instance list output for instances running on the wrong node
514 - a few fixes to the cluster-merge tool, but it still cannot merge
515 multi-node groups (currently it is not recommended to use this tool)
521 - Improved network configuration for the KVM hypervisor
522 - Added e1000 as a supported NIC for Xen-HVM
523 - Improved the lvmstrap tool to also be able to use partitions, as
524 opposed to full disks
525 - Improved speed of disk wiping (the cluster parameter
526 ``prealloc_wipe_disks``, so that it has a low impact on the total time
527 of instance creations
528 - Added documentation for the OS parameters
529 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
530 hypervisor is not responding
531 - Added display of blacklisted and hidden OS information in
533 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
534 NIC and node parameters, which might create problems with currently
535 invalid (but undetected) configuration files, but prevents validation
536 failures when unrelated parameters are modified
537 - Changed cluster initialisation to wait for the master daemon to become
539 - Expanded the RAPI interface:
541 - Added config redistribution resource
542 - Added activation/deactivation of instance disks
543 - Added export of console information
545 - Implemented log file reopening on SIGHUP, which allows using
546 logrotate(8) for the Ganeti log files
547 - Added a basic OOB helper script as an example
553 *(Released Fri, 14 Jan 2011)*
558 - Fixed timezone issues when formatting timestamps
559 - Added support for node groups, available via ``gnt-group`` and other
561 - Added out-of-band framework and management, see :doc:`design
562 document <design-oob>`
563 - Removed support for roman numbers from ``gnt-node list`` and
564 ``gnt-instance list``.
565 - Allowed modification of master network interface via ``gnt-cluster
566 modify --master-netdev``
567 - Accept offline secondaries while shutting down instance disks
568 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
569 - Added support for multiple LVM volume groups
570 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
572 - Added commands to list available fields:
574 - ``gnt-node list-fields``
575 - ``gnt-group list-fields``
576 - ``gnt-instance list-fields``
578 - Updated documentation and man pages
583 - Moved ``rapi_users`` file into separate directory, now named
584 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
586 - Added new tool for running commands on many machines,
587 ``tools/ganeti-listrunner``
588 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
589 improving the ``gnt-instance console`` output
590 - Allowed customisation of disk index separator at ``configure`` time
591 - Export node group allocation policy to :doc:`iallocator <iallocator>`
592 - Added support for non-partitioned md disks in ``lvmstrap``
593 - Added script to gracefully power off KVM instances
594 - Split ``utils`` module into smaller parts
595 - Changed query operations to return more detailed information, e.g.
596 whether an information is unavailable due to an offline node. To use
597 this new functionality, the LUXI call ``Query`` must be used. Field
598 information is now stored by the master daemon and can be retrieved
599 using ``QueryFields``. Instances, nodes and groups can also be queried
600 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
601 exposed via RAPI). The following commands make use of this
602 infrastructure change:
605 - ``gnt-group list-fields``
607 - ``gnt-node list-fields``
608 - ``gnt-instance list``
609 - ``gnt-instance list-fields``
610 - ``gnt-debug locks``
615 - New RAPI resources (see :doc:`rapi`):
619 - ``/2/groups/[group_name]``
620 - ``/2/groups/[group_name]/assign-nodes``
621 - ``/2/groups/[group_name]/modify``
622 - ``/2/groups/[group_name]/rename``
623 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
627 - Implemented ``no_install`` for instance creation
628 - Implemented OS parameters for instance reinstallation, allowing
629 use of special settings on reinstallation (e.g. for preserving data)
634 - Added IPv6 support in import/export
635 - Pause DRBD synchronization while wiping disks on instance creation
636 - Updated unittests and QA scripts
637 - Improved network parameters passed to KVM
638 - Converted man pages from docbook to reStructuredText
644 *(Released Mon, 20 Dec 2010)*
646 Released version 2.3.1~rc1 without any changes.
652 *(Released Wed, 1 Dec 2010)*
654 - impexpd: Disable OpenSSL compression in socat if possible (backport
655 from master, commit e90739d625b, see :doc:`installation guide
656 <install-quick>` for details)
657 - Changed unittest coverage report to exclude test scripts
658 - Added script to check version format
664 *(Released Wed, 1 Dec 2010)*
666 Released version 2.3.0~rc1 without any changes.
672 *(Released Fri, 19 Nov 2010)*
674 A number of bugfixes and documentation updates:
676 - Update ganeti-os-interface documentation
677 - Fixed a bug related to duplicate MACs or similar items which should be
679 - Fix breakage in OS state modify
680 - Reinstall instance: disallow offline secondaries (fixes bug related to
681 OS changing but reinstall failing)
682 - plus all the other fixes between 2.2.1 and 2.2.2
688 *(Released Tue, 2 Nov 2010)*
690 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
691 - Fixed master failover race with watcher
692 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
694 - Accept previously stopped instance for export with instance removal
695 - Simplify and extend the environment variables for instance OS scripts
696 - Added new node flags, ``master_capable`` and ``vm_capable``
697 - Added optional instance disk wiping prior during allocation. This is a
698 cluster-wide option and can be set/modified using
699 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
700 - Added IPv6 support, see :doc:`design document <design-2.3>` and
702 - Added a new watcher option (``--ignore-pause``)
703 - Added option to ignore offline node on instance start/stop
704 (``--ignore-offline``)
705 - Allow overriding OS parameters with ``gnt-instance reinstall``
706 - Added ability to change node's secondary IP address using ``gnt-node
708 - Implemented privilege separation for all daemons except
709 ``ganeti-noded``, see ``configure`` options
710 - Complain if an instance's disk is marked faulty in ``gnt-cluster
712 - Implemented job priorities (see ``ganeti(7)`` manpage)
713 - Ignore failures while shutting down instances during failover from
715 - Exit daemon's bootstrap process only once daemon is ready
716 - Export more information via ``LUInstanceQuery``/remote API
717 - Improved documentation, QA and unittests
718 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
719 restart if the file was created or changed
720 - Added LUXI protocol version sent with each request and response,
721 allowing detection of server/client mismatches
722 - Moved the Python scripts among gnt-* and ganeti-* into modules
723 - Moved all code related to setting up SSH to an external script,
725 - Infrastructure changes for node group support in future versions
731 *(Released Fri, 19 Nov 2010)*
733 A few small bugs fixed, and some improvements to the build system:
735 - Fix documentation regarding conversion to drbd
736 - Fix validation of parameters in cluster modify (``gnt-cluster modify
738 - Fix error handling in node modify with multiple changes
739 - Allow remote imports without checked names
745 *(Released Tue, 19 Oct 2010)*
747 - Disable SSL session ID cache in RPC client
753 *(Released Thu, 14 Oct 2010)*
755 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
756 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
758 - Fix problems with interaction between Curl and Python's HTTP server,
759 resulting in increased speed in many RPC calls
760 - Improve our release script to prevent breakage with older aclocal and
767 *(Released Thu, 7 Oct 2010)*
769 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
770 value retrieved from master
771 - Added support for blacklisted or hidden OS definitions
772 - Added simple lock monitor (accessible via (``gnt-debug locks``)
773 - Added support for -mem-path in KVM hypervisor abstraction layer
774 - Allow overriding instance parameters in tool for inter-cluster
775 instance moves (``tools/move-instance``)
776 - Improved opcode summaries (e.g. in ``gnt-job list``)
777 - Improve consistency of OS listing by sorting it
778 - Documentation updates
784 *(Released Fri, 8 Oct 2010)*
786 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
792 *(Released Mon, 4 Oct 2010)*
794 - Fixed regression in ``gnt-instance rename``
800 *(Released Wed, 22 Sep 2010)*
802 - Fixed OS_VARIANT variable for OS scripts
803 - Fixed cluster tag operations via RAPI
804 - Made ``setup-ssh`` exit with non-zero code if an error occurred
805 - Disabled RAPI CA checks in watcher
811 *(Released Mon, 23 Aug 2010)*
813 - Support DRBD versions of the format "a.b.c.d"
815 - Re-introduce support for usage from multiple threads in RAPI client
816 - Instance renames and modify via RAPI
817 - Work around race condition between processing and archival in job
819 - Mark opcodes following failed one as failed, too
820 - Job field ``lock_status`` was removed due to difficulties making it
821 work with the changed job queue in Ganeti 2.2; a better way to monitor
822 locks is expected for a later 2.2.x release
823 - Fixed dry-run behaviour with many commands
824 - Support ``ssh-agent`` again when adding nodes
825 - Many additional bugfixes
831 *(Released Fri, 30 Jul 2010)*
833 Important change: the internal RPC mechanism between Ganeti nodes has
834 changed from using a home-grown http library (based on the Python base
835 libraries) to use the PycURL library. This requires that PycURL is
836 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
837 against GnuTLS by default. cURL's support for GnuTLS had known issues
838 before cURL 7.21.0 and we recommend using the latest cURL release or
839 linking against OpenSSL. Most other distributions already link PycURL
840 and cURL against OpenSSL. The command::
842 python -c 'import pycurl; print pycurl.version'
844 can be used to determine the libraries PycURL and cURL are linked
847 Other significant changes:
849 - Rewrote much of the internals of the job queue, in order to achieve
850 better parallelism; this decouples job query operations from the job
851 processing, and it should allow much nicer behaviour of the master
852 daemon under load, and it also has uncovered some long-standing bugs
853 related to the job serialisation (now fixed)
854 - Added a default iallocator setting to the cluster parameters,
855 eliminating the need to always pass nodes or an iallocator for
856 operations that require selection of new node(s)
857 - Added experimental support for the LXC virtualization method
858 - Added support for OS parameters, which allows the installation of
859 instances to pass parameter to OS scripts in order to customise the
861 - Added a hypervisor parameter controlling the migration type (live or
862 non-live), since hypervisors have various levels of reliability; this
863 has renamed the 'live' parameter to 'mode'
864 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
865 logical volumes, meaning that cluster verify will ignore them and not
866 flag their presence as errors
867 - The watcher will now reset the error count for failed instances after
868 8 hours, thus allowing self-healing if the problem that caused the
869 instances to be down/fail to start has cleared in the meantime
870 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
871 check for, and warn, if the drbd module parameter ``usermode_helper``
872 is not consistent with the cluster-wide setting; this is needed to
873 make diagnose easier of failed drbd creations
874 - Started adding base IPv6 support, but this is not yet
875 enabled/available for use
876 - Rename operations (cluster, instance) will now return the new name,
877 which is especially useful if a short name was passed in
878 - Added support for instance migration in RAPI
879 - Added a tool to pre-configure nodes for the SSH setup, before joining
880 them to the cluster; this will allow in the future a simplified model
881 for node joining (but not yet fully enabled in 2.2); this needs the
882 paramiko python library
883 - Fixed handling of name-resolving errors
884 - Fixed consistency of job results on the error path
885 - Fixed master-failover race condition when executed multiple times in
887 - Fixed many bugs related to the job queue (mostly introduced during the
888 2.2 development cycle, so not all are impacting 2.1)
889 - Fixed instance migration with missing disk symlinks
890 - Fixed handling of unknown jobs in ``gnt-job archive``
891 - And many other small fixes/improvements
895 - Enhanced both the unittest and the QA coverage
896 - Switched the opcode validation to a generic model, and extended the
897 validation to all opcode parameters
898 - Changed more parts of the code that write shell scripts to use the
900 - Switched the master daemon to use the asyncore library for the Luxi
907 *(Released Thu, 17 Jun 2010)*
909 - Added tool (``move-instance``) and infrastructure to move instances
910 between separate clusters (see :doc:`separate documentation
911 <move-instance>` and :doc:`design document <design-2.2>`)
912 - Added per-request RPC timeout
913 - RAPI now requires a Content-Type header for requests with a body (e.g.
914 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
915 :rfc:`2616` (HTTP/1.1), section 7.2.1)
916 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
918 - Implemented initial support for running Ganeti daemons as separate
919 users, see configure-time flags ``--with-user-prefix`` and
920 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
922 - Instances can be removed after export (``gnt-backup export
924 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
925 key (instead of 1024 bit)
926 - Added new cluster configuration file for cluster domain secret
927 - Import/export now use SSL instead of SSH
928 - Added support for showing estimated time when exporting an instance,
929 see the ``ganeti-os-interface(7)`` manpage and look for
936 *(Released Tue, 16 Nov 2010)*
938 Some more bugfixes. Unless critical bugs occur, this will be the last
941 - Fix case of MAC special-values
942 - Fix mac checker regex
943 - backend: Fix typo causing "out of range" error
944 - Add missing --units in gnt-instance list man page
950 *(Released Tue, 24 Aug 2010)*
953 - Don't ignore secondary node silently on non-mirrored disk templates
955 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
956 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
957 - Properly document the usb_mouse parameter
958 - Fix path in ganeti-rapi(8) (issue 116)
959 - Adjust error message when the ganeti user's .ssh directory is
961 - Add same-node-check when changing the disk template to drbd
967 *(Released Fri, 16 Jul 2010)*
970 - Add an option to only select some reboot types during qa/burnin.
971 (on some hypervisors consequent reboots are not supported)
972 - Fix infrequent race condition in master failover. Sometimes the old
973 master ip address would be still detected as up for a short time
974 after it was removed, causing failover to fail.
975 - Decrease mlockall warnings when the ctypes module is missing. On
976 Python 2.4 we support running even if no ctypes module is installed,
977 but we were too verbose about this issue.
978 - Fix building on old distributions, on which man doesn't have a
980 - Fix RAPI not to ignore the MAC address on instance creation
981 - Implement the old instance creation format in the RAPI client.
987 *(Released Thu, 01 Jul 2010)*
989 A small bugfix release:
990 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
991 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
992 a non-existing option
993 - Fix instance create: the --force-variant option was ignored
994 - Improve pylint 0.21 compatibility and warnings with Python 2.6
995 - Fix modify node storage with non-FQDN arguments
996 - Fix RAPI client to authenticate under Python 2.6 when used
997 for more than 5 requests needing authentication
998 - Fix gnt-instance modify -t (storage) giving a wrong error message
999 when converting a non-shutdown drbd instance to plain
1005 *(Released Fri, 18 Jun 2010)*
1007 A small bugfix release:
1009 - Fix live migration of KVM instances started with older Ganeti
1010 versions which had fewer hypervisor parameters
1011 - Fix gnt-instance grow-disk on down instances
1012 - Fix an error-reporting bug during instance migration
1013 - Better checking of the ``--net`` and ``--disk`` values, to avoid
1014 silently ignoring broken ones
1015 - Fix an RPC error reporting bug affecting, for example, RAPI client
1017 - Fix bug triggered by different API version os-es on different nodes
1018 - Fix a bug in instance startup with custom hvparams: OS level
1019 parameters would fail to be applied.
1020 - Fix the RAPI client under Python 2.6 (but more work is needed to
1021 make it work completely well with OpenSSL)
1022 - Fix handling of errors when resolving names from DNS
1028 *(Released Thu, 3 Jun 2010)*
1030 A medium sized development cycle. Some new features, and some
1031 fixes/small improvements/cleanups.
1033 Significant features
1034 ~~~~~~~~~~~~~~~~~~~~
1036 The node deamon now tries to mlock itself into memory, unless the
1037 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1038 its logs, and falls back to console logging. This allows emergency
1039 features such as ``gnt-node powercycle`` to work even in the event of a
1040 broken node disk (tested offlining the disk hosting the node's
1041 filesystem and dropping its memory caches; don't try this at home)
1043 KVM: add vhost-net acceleration support. It can be tested with a new
1044 enough version of the kernel and of qemu-kvm.
1046 KVM: Add instance chrooting feature. If you use privilege dropping for
1047 your VMs you can also now force them to chroot to an empty directory,
1048 before starting the emulated guest.
1050 KVM: Add maximum migration bandwith and maximum downtime tweaking
1051 support (requires a new-enough version of qemu-kvm).
1053 Cluster verify will now warn if the master node doesn't have the master
1054 ip configured on it.
1056 Add a new (incompatible) instance creation request format to RAPI which
1057 supports all parameters (previously only a subset was supported, and it
1058 wasn't possible to extend the old format to accomodate all the new
1059 features. The old format is still supported, and a client can check for
1060 this feature, before using it, by checking for its presence in the
1061 ``features`` RAPI resource.
1063 Now with ancient latin support. Try it passing the ``--roman`` option to
1064 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1065 (requires the python-roman module to be installed, in order to work).
1070 As usual many internal code refactorings, documentation updates, and
1073 - Lots of improvements and cleanups to the experimental Remote API
1074 (RAPI) client library.
1075 - A new unit test suite for the core daemon libraries.
1076 - A fix to creating missing directories makes sure the umask is not
1077 applied anymore. This enforces the same directory permissions
1079 - Better handling terminating daemons with ctrl+c (used when running
1080 them in debugging mode).
1081 - Fix a race condition in live migrating a KVM instance, when stat()
1082 on the old proc status file returned EINVAL, which is an unexpected
1084 - Fixed manpage checking with newer man and utf-8 charachters. But now
1085 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1091 *(Released Fri, 7 May 2010)*
1093 Fix a bug which prevented untagged KVM instances from starting.
1099 *(Released Fri, 7 May 2010)*
1101 Another release with a long development cycle, during which many
1102 different features were added.
1104 Significant features
1105 ~~~~~~~~~~~~~~~~~~~~
1107 The KVM hypervisor now can run the individual instances as non-root, to
1108 reduce the impact of a VM being hijacked due to bugs in the
1109 hypervisor. It is possible to run all instances as a single (non-root)
1110 user, to manually specify a user for each instance, or to dynamically
1111 allocate a user out of a cluster-wide pool to each instance, with the
1112 guarantee that no two instances will run under the same user ID on any
1115 An experimental RAPI client library, that can be used standalone
1116 (without the other Ganeti libraries), is provided in the source tree as
1117 ``lib/rapi/client.py``. Note this client might change its interface in
1118 the future, as we iterate on its capabilities.
1120 A new command, ``gnt-cluster renew-crypto`` has been added to easily
1121 replace the cluster's certificates and crypto keys. This might help in
1122 case they have been compromised, or have simply expired.
1124 A new disk option for instance creation has been added that allows one
1125 to "adopt" currently existing logical volumes, with data
1126 preservation. This should allow easier migration to Ganeti from
1127 unmanaged (or managed via other software) instances.
1129 Another disk improvement is the possibility to convert between redundant
1130 (DRBD) and plain (LVM) disk configuration for an instance. This should
1131 allow better scalability (starting with one node and growing the
1132 cluster, or shrinking a two-node cluster to one node).
1134 A new feature that could help with automated node failovers has been
1135 implemented: if a node sees itself as offline (by querying the master
1136 candidates), it will try to shutdown (hard) all instances and any active
1137 DRBD devices. This reduces the risk of duplicate instances if an
1138 external script automatically failovers the instances on such nodes. To
1139 enable this, the cluster parameter ``maintain_node_health`` should be
1140 enabled; in the future this option (per the name) will enable other
1141 automatic maintenance features.
1143 Instance export/import now will reuse the original instance
1144 specifications for all parameters; that means exporting an instance,
1145 deleting it and the importing it back should give an almost identical
1146 instance. Note that the default import behaviour has changed from
1147 before, where it created only one NIC; now it recreates the original
1150 Cluster verify has added a few new checks: SSL certificates validity,
1151 /etc/hosts consistency across the cluster, etc.
1156 As usual, many internal changes were done, documentation fixes,
1159 - Fixed cluster initialization with disabled cluster storage (regression
1160 introduced in 2.1.1)
1161 - File-based storage supports growing the disks
1162 - Fixed behaviour of node role changes
1163 - Fixed cluster verify for some corner cases, plus a general rewrite of
1164 cluster verify to allow future extension with more checks
1165 - Fixed log spamming by watcher and node daemon (regression introduced
1167 - Fixed possible validation issues when changing the list of enabled
1169 - Fixed cleanup of /etc/hosts during node removal
1170 - Fixed RAPI response for invalid methods
1171 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1172 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
1173 from ide disks, etc.)
1174 - Allow OS changes without re-installation (to record a changed OS
1175 outside of Ganeti, or to allow OS renames)
1176 - Allow instance creation without OS installation (useful for example if
1177 the OS will be installed manually, or restored from a backup not in
1179 - Implemented option to make cluster ``copyfile`` use the replication
1181 - Added list of enabled hypervisors to ssconf (possibly useful for
1183 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1185 - A partial form of node re-IP is possible via node readd, which now
1186 allows changed node primary IP
1187 - Command line utilities now show an informational message if the job is
1189 - The logs of the master daemon now show the PID/UID/GID of the
1196 *(Released Fri, 12 Mar 2010)*
1198 During the 2.1.0 long release candidate cycle, a lot of improvements and
1199 changes have accumulated with were released later as 2.1.1.
1204 The node evacuate command (``gnt-node evacuate``) was significantly
1205 rewritten, and as such the IAllocator protocol was changed - a new
1206 request type has been added. This unfortunate change during a stable
1207 series is designed to improve performance of node evacuations; on
1208 clusters with more than about five nodes and which are well-balanced,
1209 evacuation should proceed in parallel for all instances of the node
1210 being evacuated. As such, any existing IAllocator scripts need to be
1211 updated, otherwise the above command will fail due to the unknown
1212 request. The provided "dumb" allocator has not been updated; but the
1213 ganeti-htools package supports the new protocol since version 0.2.4.
1215 Another important change is increased validation of node and instance
1216 names. This might create problems in special cases, if invalid host
1217 names are being used.
1219 Also, a new layer of hypervisor parameters has been added, that sits at
1220 OS level between the cluster defaults and the instance ones. This allows
1221 customisation of virtualization parameters depending on the installed
1222 OS. For example instances with OS 'X' may have a different KVM kernel
1223 (or any other parameter) than the cluster defaults. This is intended to
1224 help managing a multiple OSes on the same cluster, without manual
1225 modification of each instance's parameters.
1227 A tool for merging clusters, ``cluster-merge``, has been added in the
1228 tools sub-directory.
1233 - Improved the int/float conversions that should make the code more
1234 robust in face of errors from the node daemons
1235 - Fixed the remove node code in case of internal configuration errors
1236 - Fixed the node daemon behaviour in face of inconsistent queue
1237 directory (e.g. read-only file-system where we can't open the files
1239 - Fixed the behaviour of gnt-node modify for master candidate demotion;
1240 now it either aborts cleanly or, if given the new "auto_promote"
1241 parameter, will automatically promote other nodes as needed
1242 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1243 completely prevent Ganeti from working
1244 - Fixed bug for instance export when not all disks were successfully
1246 - Fixed behaviour of node add when the new node is slow in starting up
1248 - Fixed handling of signals in the LUXI client, which should improve
1249 behaviour of command-line scripts
1250 - Added checks for invalid node/instance names in the configuration (now
1251 flagged during cluster verify)
1252 - Fixed watcher behaviour for disk activation errors
1253 - Fixed two potentially endless loops in http library, which led to the
1254 RAPI daemon hanging and consuming 100% CPU in some cases
1255 - Fixed bug in RAPI daemon related to hashed passwords
1256 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1258 - Enhanced compatibility with non-Debian OSes, but not using absolute
1259 path in some commands and allowing customisation of the ssh
1260 configuration directory
1261 - Fixed possible future issue with new Python versions by abiding to the
1262 proper use of ``__slots__`` attribute on classes
1263 - Added checks that should prevent directory traversal attacks
1264 - Many documentation fixes based on feedback from users
1269 - Added an "early_release" more for instance replace disks and node
1270 evacuate, where we release locks earlier and thus allow higher
1271 parallelism within the cluster
1272 - Added watcher hooks, intended to allow the watcher to restart other
1273 daemons (e.g. from the ganeti-nbma project), but they can be used of
1274 course for any other purpose
1275 - Added a compile-time disable for DRBD barriers, to increase
1276 performance if the administrator trusts the power supply or the
1277 storage system to not lose writes
1278 - Added the option of using syslog for logging instead of, or in
1279 addition to, Ganeti's own log files
1280 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1281 can indeed boot from a paravirtual NIC
1282 - Added a generic debug level for many operations; while this is not
1283 used widely yet, it allows one to pass the debug value all the way to
1285 - Enhanced the hooks environment for instance moves (failovers,
1286 migrations) where the primary/secondary nodes changed during the
1287 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1288 - Enhanced data validations for many user-supplied values; one important
1289 item is the restrictions imposed on instance and node names, which
1290 might reject some (invalid) host names
1291 - Add a configure-time option to disable file-based storage, if it's not
1292 needed; this allows greater security separation between the master
1293 node and the other nodes from the point of view of the inter-node RPC
1295 - Added user notification in interactive tools if job is waiting in the
1296 job queue or trying to acquire locks
1297 - Added log messages when a job is waiting for locks
1298 - Added filtering by node tags in instance operations which admit
1299 multiple instances (start, stop, reboot, reinstall)
1300 - Added a new tool for cluster mergers, ``cluster-merge``
1301 - Parameters from command line which are of the form ``a=b,c=d`` can now
1302 use backslash escapes to pass in values which contain commas,
1303 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1305 - For KVM, the instance name is the first parameter passed to KVM, so
1306 that it's more visible in the process list
1312 *(Released Tue, 2 Mar 2010)*
1314 Ganeti 2.1 brings many improvements with it. Major changes:
1316 - Added infrastructure to ease automated disk repairs
1317 - Added new daemon to export configuration data in a cheaper way than
1318 using the remote API
1319 - Instance NICs can now be routed instead of being associated with a
1321 - Improved job locking logic to reduce impact of jobs acquiring multiple
1322 locks waiting for other long-running jobs
1324 In-depth implementation details can be found in the Ganeti 2.1 design
1330 - Added chroot hypervisor
1331 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1333 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1334 ``bootloader_path`` and ``bootloader_args``)
1335 - Added the ``use_localtime`` option for the xen-hvm and kvm
1336 hypervisors, and the default value for this has changed to false (in
1337 2.0 xen-hvm always enabled it)
1338 - Added luxi call to submit multiple jobs in one go
1339 - Added cluster initialization option to not modify ``/etc/hosts``
1341 - Added network interface parameters
1342 - Added dry run mode to some LUs
1343 - Added RAPI resources:
1345 - ``/2/instances/[instance_name]/info``
1346 - ``/2/instances/[instance_name]/replace-disks``
1347 - ``/2/nodes/[node_name]/evacuate``
1348 - ``/2/nodes/[node_name]/migrate``
1349 - ``/2/nodes/[node_name]/role``
1350 - ``/2/nodes/[node_name]/storage``
1351 - ``/2/nodes/[node_name]/storage/modify``
1352 - ``/2/nodes/[node_name]/storage/repair``
1354 - Added OpCodes to evacuate or migrate all instances on a node
1355 - Added new command to list storage elements on nodes (``gnt-node
1356 list-storage``) and modify them (``gnt-node modify-storage``)
1357 - Added new ssconf files with master candidate IP address
1358 (``ssconf_master_candidates_ips``), node primary IP address
1359 (``ssconf_node_primary_ips``) and node secondary IP address
1360 (``ssconf_node_secondary_ips``)
1361 - Added ``ganeti-confd`` and a client library to query the Ganeti
1362 configuration via UDP
1363 - Added ability to run hooks after cluster initialization and before
1365 - Added automatic mode for disk replace (``gnt-instance replace-disks
1367 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1368 after catastrophic data-loss
1369 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1371 - Added ``gnt-instance move`` command to move instances
1372 - Added ``gnt-cluster watcher`` command to control watcher
1373 - Added ``gnt-node powercycle`` command to powercycle nodes
1374 - Added new job status field ``lock_status``
1375 - Added parseable error codes to cluster verification (``gnt-cluster
1376 verify --error-codes``) and made output less verbose (use
1377 ``--verbose`` to restore previous behaviour)
1378 - Added UUIDs to the main config entities (cluster, nodes, instances)
1379 - Added support for OS variants
1380 - Added support for hashed passwords in the Ganeti remote API users file
1382 - Added option to specify maximum timeout on instance shutdown
1383 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1384 - Added new helper script to start and stop Ganeti daemons
1385 (``daemon-util``), with the intent to reduce the work necessary to
1386 adjust Ganeti for non-Debian distributions and to start/stop daemons
1388 - Added more unittests
1389 - Fixed critical bug in ganeti-masterd startup
1390 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1391 now customisable at the cluster level for both the KVM and Xen
1392 hypervisors using the new ``migration_port`` parameter
1393 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1394 reinstalling an instance
1395 - Allowed ``@`` in tag names
1396 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1397 - Many documentation updates
1398 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1399 - ``gnt-instance reinstall`` can now reinstall multiple instances
1400 - Updated many command line parameters
1401 - Introduced new OS API version 15
1402 - No longer support a default hypervisor
1403 - Treat virtual LVs as inexistent
1404 - Improved job locking logic to reduce lock contention
1405 - Match instance and node names case insensitively
1406 - Reimplemented bash completion script to be more complete
1413 *(Released Thu, 4 Feb 2010)*
1415 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1416 - Fix a string formatting bug
1417 - Improve safety of the code in some error paths
1418 - Improve data validation in the master of values returned from nodes
1424 *(Released Thu, 17 Dec 2009)*
1426 - Fix security issue due to missing validation of iallocator names; this
1427 allows local and remote execution of arbitrary executables
1428 - Fix failure of gnt-node list during instance removal
1429 - Ship the RAPI documentation in the archive
1435 *(Released Wed, 30 Sep 2009)*
1437 - Fixed many wrong messages
1438 - Fixed a few bugs related to the locking library
1439 - Fixed MAC checking at instance creation time
1440 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1441 - Fixed a few issues related to signal handling in both daemons and
1443 - Fixed the example startup script provided
1444 - Fixed insserv dependencies in the example startup script (patch from
1446 - Fixed handling of drained nodes in the iallocator framework
1447 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1449 - Fixed error related to invalid job IDs in job polling
1450 - Fixed job/opcode persistence on unclean master shutdown
1451 - Fixed handling of partial job processing after unclean master
1453 - Fixed error reporting from LUs, previously all errors were converted
1454 into execution errors
1455 - Fixed error reporting from burnin
1456 - Decreased significantly the memory usage of the job queue
1457 - Optimised slightly multi-job submission
1458 - Optimised slightly opcode loading
1459 - Backported the multi-job submit framework from the development
1460 branch; multi-instance start and stop should be faster
1461 - Added script to clean archived jobs after 21 days; this will reduce
1462 the size of the queue directory
1463 - Added some extra checks in disk size tracking
1464 - Added an example ethers hook script
1465 - Added a cluster parameter that prevents Ganeti from modifying of
1467 - Added more node information to RAPI responses
1468 - Added a ``gnt-job watch`` command that allows following the ouput of a
1470 - Added a bind-address option to ganeti-rapi
1471 - Added more checks to the configuration verify
1472 - Enhanced the burnin script such that some operations can be retried
1474 - Converted instance reinstall to multi-instance model
1480 *(Released Fri, 7 Aug 2009)*
1482 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1483 to allow using the pre-2.0.2 behaviour in activation, if any existing
1484 instances have mismatched disk sizes in the configuration
1485 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1486 any configuration mismatches for disk sizes
1487 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1488 failover to work on two-node clusters
1489 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1491 - Fixed detection of OS script errors in ``gnt-backup export``
1492 - Fixed exit code of ``gnt-backup export``
1498 *(Released Fri, 17 Jul 2009)*
1500 - Added experimental support for stripped logical volumes; this should
1501 enhance performance but comes with a higher complexity in the block
1502 device handling; stripping is only enabled when passing
1503 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1504 affected even in the non-stripped mode
1505 - Improved resiliency against transient failures at the end of DRBD
1506 resyncs, and in general of DRBD resync checks
1507 - Fixed a couple of issues with exports and snapshot errors
1508 - Fixed a couple of issues in instance listing
1509 - Added display of the disk size in ``gnt-instance info``
1510 - Fixed checking for valid OSes in instance creation
1511 - Fixed handling of the "vcpus" parameter in instance listing and in
1512 general of invalid parameters
1513 - Fixed http server library, and thus RAPI, to handle invalid
1514 username/password combinations correctly; this means that now they
1515 report unauthorized for queries too, not only for modifications,
1516 allowing earlier detect of configuration problems
1517 - Added a new "role" node list field, equivalent to the master/master
1518 candidate/drained/offline flags combinations
1519 - Fixed cluster modify and changes of candidate pool size
1520 - Fixed cluster verify error messages for wrong files on regular nodes
1521 - Fixed a couple of issues with node demotion from master candidate role
1522 - Fixed node readd issues
1523 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1524 - Added a new ``--no-voting`` option for masterfailover to fix failover
1525 on two-nodes clusters when the former master node is unreachable
1526 - Added instance reinstall over RAPI
1532 *(Released Tue, 16 Jun 2009)*
1534 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1535 allow re-adding the start in single-user option (regression from 1.2)
1536 - the watcher writes the instance status to a file, to allow monitoring
1537 to report the instance status (from the master) based on cached
1538 results of the watcher's queries; while this can get stale if the
1539 watcher is being locked due to other work on the cluster, this is
1540 still an improvement
1541 - the watcher now also restarts the node daemon and the rapi daemon if
1543 - fixed the watcher to handle full and drained queue cases
1544 - hooks export more instance data in the environment, which helps if
1545 hook scripts need to take action based on the instance's properties
1546 (no longer need to query back into ganeti)
1547 - instance failovers when the instance is stopped do not check for free
1548 RAM, so that failing over a stopped instance is possible in low memory
1550 - rapi uses queries for tags instead of jobs (for less job traffic), and
1551 for cluster tags it won't talk to masterd at all but read them from
1553 - a couple of error handling fixes in RAPI
1554 - drbd handling: improved the error handling of inconsistent disks after
1555 resync to reduce the frequency of "there are some degraded disks for
1556 this instance" messages
1557 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1558 error handling path called a wrong function name)
1564 *(Released Wed, 27 May 2009)*
1566 - no changes from rc5
1569 Version 2.0 release candidate 5
1570 -------------------------------
1572 *(Released Wed, 20 May 2009)*
1574 - fix a couple of bugs (validation, argument checks)
1575 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1576 - some small improvements to RAPI and IAllocator
1577 - make watcher automatically start the master daemon if down
1580 Version 2.0 release candidate 4
1581 -------------------------------
1583 *(Released Mon, 27 Apr 2009)*
1585 - change the OS list to not require locks; this helps with big clusters
1586 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1587 volume group is broken
1588 - ``gnt-instance info``, without any arguments, doesn't run for all
1589 instances anymore; either pass ``--all`` or pass the desired
1590 instances; this helps against mistakes on big clusters where listing
1591 the information for all instances takes a long time
1592 - miscellaneous doc and man pages fixes
1595 Version 2.0 release candidate 3
1596 -------------------------------
1598 *(Released Wed, 8 Apr 2009)*
1600 - Change the internal locking model of some ``gnt-node`` commands, in
1601 order to reduce contention (and blocking of master daemon) when
1602 batching many creation/reinstall jobs
1603 - Fixes to Xen soft reboot
1604 - No longer build documentation at build time, instead distribute it in
1605 the archive, in order to reduce the need for the whole docbook/rst
1609 Version 2.0 release candidate 2
1610 -------------------------------
1612 *(Released Fri, 27 Mar 2009)*
1614 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1615 - Fix watcher startup sequence, improves the behaviour of busy clusters
1616 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1617 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1618 bind address and other places
1619 - Some documentation fixes and updates
1622 Version 2.0 release candidate 1
1623 -------------------------------
1625 *(Released Mon, 2 Mar 2009)*
1627 - More documentation updates, now all docs should be more-or-less
1629 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1631 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1633 - Fix an issue related to $libdir/run/ganeti and cluster creation
1639 *(Released Thu, 19 Feb 2009)*
1641 - Xen PVM and KVM have switched the default value for the instance root
1642 disk to the first partition on the first drive, instead of the whole
1643 drive; this means that the OS installation scripts must be changed
1645 - Man pages have been updated
1646 - RAPI has been switched by default to HTTPS, and the exported functions
1647 should all work correctly
1648 - RAPI v1 has been removed
1649 - Many improvements to the KVM hypervisor
1650 - Block device errors are now better reported
1651 - Many other bugfixes and small improvements
1657 *(Released Mon, 26 Jan 2009)*
1659 - Version 2 is a general rewrite of the code and therefore the
1660 differences are too many to list, see the design document for 2.0 in
1661 the ``doc/`` subdirectory for more details
1662 - In this beta version there is not yet a migration path from 1.2 (there
1663 will be one in the final 2.0 release)
1664 - A few significant changes are:
1666 - all commands are executed by a daemon (``ganeti-masterd``) and the
1667 various ``gnt-*`` commands are just front-ends to it
1668 - all the commands are entered into, and executed from a job queue,
1669 see the ``gnt-job(8)`` manpage
1670 - the RAPI daemon supports read-write operations, secured by basic
1671 HTTP authentication on top of HTTPS
1672 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1673 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1674 to migrate to DRBD 8 first while still running Ganeti 1.2)
1675 - DRBD devices are using statically allocated minor numbers, which
1676 will be assigned to existing instances during the migration process
1677 - there is support for both Xen PVM and Xen HVM instances running on
1679 - KVM virtualization is supported too
1680 - file-based storage has been implemented, which means that it is
1681 possible to run the cluster without LVM and DRBD storage, for
1682 example using a shared filesystem exported from shared storage (and
1683 still have live migration)
1689 *(Released Tue, 13 Jan 2009)*
1691 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1692 - Reuse the old instance mac address by default on instance import, if
1693 the instance name is the same.
1694 - Handle situations in which the node info rpc returns incomplete
1696 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1697 - Improved version of batcher:
1699 - state file support
1700 - instance mac address support
1701 - support for HVM clusters/instances
1703 - Add an option to show the number of cpu sockets and nodes in
1705 - Support OSes that handle more than one version of the OS api (but do
1706 not change the current API in any other way)
1707 - Fix ``gnt-node migrate``
1708 - ``gnt-debug`` man page
1709 - Fixes various more typos and small issues
1710 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1716 *(Released Wed, 24 Sep 2008)*
1718 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1719 type of disk exported to fully virtualized instances.
1720 - provide access to the serial console of HVM instances
1721 - instance auto_balance flag, set by default. If turned off it will
1722 avoid warnings on cluster verify if there is not enough memory to fail
1723 over an instance. in the future it will prevent automatically failing
1724 it over when we will support that.
1725 - batcher tool for instance creation, see ``tools/README.batcher``
1726 - ``gnt-instance reinstall --select-os`` to interactively select a new
1727 operating system when reinstalling an instance.
1728 - when changing the memory amount on instance modify a check has been
1729 added that the instance will be able to start. also warnings are
1730 emitted if the instance will not be able to fail over, if auto_balance
1732 - documentation fixes
1733 - sync fields between ``gnt-instance list/modify/add/import``
1734 - fix a race condition in drbd when the sync speed was set after giving
1735 the device a remote peer.
1741 *(Released Tue, 22 Jul 2008)*
1743 - note: the allowed size and number of tags per object were reduced
1744 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1745 - fixed twisted 8.x compatibility
1746 - fixed ``gnt-instance replace-disks`` with iallocator
1747 - add TCP keepalives on twisted connections to detect restarted nodes
1748 - disk increase support, see ``gnt-instance grow-disk``
1749 - implement bulk node/instance query for RAPI
1750 - add tags in node/instance listing (optional)
1751 - experimental migration (and live migration) support, read the man page
1752 for ``gnt-instance migrate``
1753 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1754 has some small improvements in handling its state file
1760 *(Released Fri, 13 Jun 2008)*
1762 - Experimental readonly, REST-based remote API implementation;
1763 automatically started on master node, TCP port 5080, if enabled by
1764 ``--enable-rapi`` parameter to configure script.
1765 - Instance allocator support. Add and import instance accept a
1766 ``--iallocator`` parameter, and call that instance allocator to decide
1767 which node to use for the instance. The iallocator document describes
1768 what's expected from an allocator script.
1769 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1770 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1771 node is lost there is still enough memory to fail over the instances
1773 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1774 ``gnt-cluster verify``, to check for site-specific compliance. All the
1775 hooks will run, and their output, if any, will be displayed. Any
1776 failing hook will make the verification return an error value.
1777 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1778 primary and secondary interfaces
1779 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1780 that is still declared as part of the cluster and has failed.
1781 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1782 specifying output fields, that just adds the chosen fields to the
1784 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1785 export from the filesystem.
1786 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1787 have been added. Using them you can enable/disable acpi and pae
1788 support, and specify a path for a cd image to be exported to the
1789 instance. These parameters as the name suggest only work on HVM
1791 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1792 PAE support will be set to the previously hardcoded values, but the
1793 (previously hardcoded) path to the CDROM ISO image will be unset and
1794 if required, needs to be set manually with ``gnt-instance modify``
1796 - The address to which an instance's VNC console is bound is now
1797 selectable per-instance, rather than being cluster wide. Of course
1798 this only applies to instances controlled via VNC, so currently just
1799 applies to HVM clusters.
1805 *(Released Mon, 18 Feb 2008)*
1807 - more tweaks to the disk activation code (especially helpful for DRBD)
1808 - change the default ``gnt-instance list`` output format, now there is
1809 one combined status field (see the manpage for the exact values this
1811 - some more fixes for the mac export to hooks change
1812 - make Ganeti not break with DRBD 8.2.x (which changed the version
1813 format in ``/proc/drbd``) (issue 24)
1814 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1815 template, allowing migration from DRBD0.7+MD to DRBD8
1821 *(Released Wed, 30 Jan 2008)*
1823 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1825 - add command aliases infrastructure and a few aliases
1826 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1827 man pages and the ``--help`` option of ``gnt-node
1828 list``/``gnt-instance list``
1829 - fix ``gnt-backup list`` with down nodes (issue 21)
1830 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1831 - fix the dist archive and add a check for including svn/git files in
1833 - some developer-related changes: improve the burnin and the QA suite,
1834 add an upload script for testing during development
1840 *(Released Wed, 16 Jan 2008)*
1842 - experimental HVM support, read the install document, section
1843 "Initializing the cluster"
1844 - allow for the PVM hypervisor per-instance kernel and initrd paths
1845 - add a new command ``gnt-cluster verify-disks`` which uses a new
1846 algorithm to improve the reconnection of the DRBD pairs if the device
1847 on the secondary node has gone away
1848 - make logical volume code auto-activate LVs at disk activation time
1849 - slightly improve the speed of activating disks
1850 - allow specification of the MAC address at instance creation time, and
1851 changing it later via ``gnt-instance modify``
1852 - fix handling of external commands that generate lots of output on
1854 - update documentation with regard to minimum version of DRBD8 supported
1860 *(Released Tue, 4 Dec 2007)*
1862 - Log the ``xm create`` output to the node daemon log on failure (to
1863 help diagnosing the error)
1864 - In debug mode, log all external commands output if failed to the logs
1865 - Change parsing of lvm commands to ignore stderr
1871 *(Released Wed, 28 Nov 2007)*
1873 - Another round of updates to the DRBD 8 code to deal with more failures
1874 in the replace secondary node operation
1875 - Some more logging of failures in disk operations (lvm, drbd)
1876 - A few documentation updates
1883 *(Released Tue, 13 Nov 2007)*
1885 - Change configuration file format from Python's Pickle to JSON.
1886 Upgrading is possible using the cfgupgrade utility.
1887 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1888 faster replace disks and is more stable (DRBD 8 has many improvements
1889 compared to DRBD 0.7)
1890 - Added command line tags support (see man pages for ``gnt-instance``,
1891 ``gnt-node``, ``gnt-cluster``)
1892 - Added instance rename support
1893 - Added multi-instance startup/shutdown
1894 - Added cluster rename support
1895 - Added ``gnt-node evacuate`` to simplify some node operations
1896 - Added instance reboot operation that can speedup reboot as compared to
1898 - Soften the requirement that hostnames are in FQDN format
1899 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1901 - Removed dependency on debian's patched fping that uses the
1902 non-standard ``-S`` option
1903 - Now the OS definitions are searched for in multiple, configurable
1904 paths (easier for distros to package)
1905 - Some changes to the hooks infrastructure (especially the new
1906 post-configuration update hook)
1907 - Other small bugfixes
1909 .. vim: set textwidth=72 syntax=rst :