10 Incompatible/important changes:
12 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
13 resource's ``ip_check`` parameter changed from ``True`` to ``False``
14 to match the underlying LUXI interface
15 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
16 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
17 not maintain backwards-compatibility as the underlying operation
18 changed in an incompatible way. The RAPI client can talk to old
19 servers, but it needs to be told so as the return value changed.
20 - When creating file-based instances via RAPI, the ``file_driver``
21 parameter no longer defaults to ``loop`` and must be specified
22 - The deprecated "bridge" nic parameter is no longer supported. Use
24 - Support for the undocumented and deprecated RAPI instance creation
25 request format version 0 has been dropped. Use version 1, supported
26 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
27 - Pyparsing 1.4.6 or above is required, see :doc:`installation
28 documentation <install>`
29 - The "cluster-verify" hooks are now executed per group by the
30 OP_CLUSTER_VERIFY_GROUP opcode. This maintains the same behavior if
31 you just run "gnt-cluster verify", which generates one op per group.
32 - The environment as passed to the OS scripts is cleared, and thus no
33 environment variables defined in the node daemon's environment will be
34 inherited by the scripts
40 *(Released Fri, 5 Aug 2011)*
42 Many bug-fixes and a few small features:
44 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
45 issues when you tried to add an instance with two MAC addresses in one
47 - KVM: fixed per-instance stored UID value
48 - KVM: configure bridged NICs at migration start
49 - KVM: Fix a bug where instance will not start with never KVM versions
51 - Added OS search path to ``gnt-cluster info``
52 - Fixed an issue with ``file_storage_dir`` where you were forced to
53 provide an absolute path, but the documentation states it is a
54 relative path, the documentation was right
55 - Added a new parameter to instance stop/start called ``--no-remember``
56 that will make the state change to not be remembered
57 - Implemented ``no_remember`` at RAPI level
58 - Improved the documentation
59 - Node evacuation: don't call IAllocator if node is already empty
60 - Fixed bug in DRBD8 replace disks on current nodes
61 - Fixed bug in recreate-disks for DRBD instances
62 - Moved assertion checking locks in ``gnt-instance replace-disks``
63 causing it to abort with not owning the right locks for some situation
64 - Job queue: Fixed potential race condition when cancelling queued jobs
65 - Fixed off-by-one bug in job serial generation
66 - ``gnt-node volumes``: Fix instance names
67 - Fixed aliases in bash completion
68 - Fixed a bug in reopening log files after beeing sent a SIGHUP
69 - Added a flag to burnin to allow specifying VCPU count
70 - Bugfixes to non-root Ganeti configuration
76 *(Released Thu, 12 May 2011)*
78 Many bug-fixes and a few new small features:
80 - Fixed a bug related to log opening failures
81 - Fixed a bug in instance listing with orphan instances
82 - Fixed a bug which prevented resetting the cluster-level node parameter
83 ``oob_program`` to the default
84 - Many fixes related to the ``cluster-merge`` tool
85 - Fixed a race condition in the lock monitor, which caused failures
86 during (at least) creation of many instances in parallel
87 - Improved output for gnt-job info
88 - Removed the quiet flag on some ssh calls which prevented debugging
90 - Improved the N+1 failure messages in cluster verify by actually
91 showing the memory values (needed and available)
92 - Increased lock attempt timeouts so that when executing long operations
93 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
94 too early and thus prevent the use of the 'fair' mechanism
95 - Changed instance query data (``gnt-instance info``) to not acquire
96 locks unless needed, thus allowing its use on locked instance if only
97 static information is asked for
98 - Improved behaviour with filesystems that do not support rename on an
100 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
101 kept locks on all nodes during the wipe, which is unneeded
102 - Fixed ``gnt-watcher`` handling of errors during hooks execution
103 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
104 10GiB) which caused the wipe to fail right at the end in some cases
105 - Fixed master IP activation when doing master failover with no-voting
106 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
107 the master node itself
108 - Fixed potential data-loss in under disk full conditions, where Ganeti
109 wouldn't check correctly the return code and would consider
110 partially-written files 'correct'
111 - Fixed bug related to multiple VGs and DRBD disk replacing
112 - Added new disk parameter ``metavg`` that allows placement of the meta
113 device for DRBD in a different volume group
114 - Fixed error handling in the node daemon when the system libc doesn't
115 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
116 - Fixed lock release during replace-disks, which kept cluster-wide locks
117 when doing disk replaces with an iallocator script
118 - Added check for missing bridges in cluster verify
119 - Handle EPIPE errors while writing to the terminal better, so that
120 piping the output to e.g. ``less`` doesn't cause a backtrace
121 - Fixed rare case where a ^C during Luxi calls could have been
122 interpreted as server errors, instead of simply terminating
123 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
125 - Added a few more parameters to the KVM hypervisor, allowing a second
126 CDROM, custom disk type for CDROMs and a floppy image
127 - Removed redundant message in instance rename when the name is given
129 - Added option to ``gnt-instance recreate-disks`` to allow creating the
130 disks on new nodes, allowing recreation when the original instance
131 nodes are completely gone
132 - Added option when converting disk templates to DRBD to skip waiting
133 for the resync, in order to make the instance available sooner
134 - Added two new variables to the OS scripts environment (containing the
136 - Made the root_path and optional parameter for the xen-pvm hypervisor,
137 to allow use of ``pvgrub`` as bootloader
138 - Changed the instance memory modifications to only check out-of-memory
139 conditions on memory increases, and turned the secondary node warnings
140 into errors (they can still be overridden via ``--force``)
141 - Fixed the handling of a corner case when the Python installation gets
142 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
143 to execute a command that doesn't exist
144 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
145 node of the instance returned failures during instance shutdown; this
146 adds the option ``--ignore-consistency`` to gnt-instance move
148 And as usual, various improvements to the error messages, documentation
154 *(Released Wed, 09 Mar 2011)*
156 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
157 the RAPI users file if run twice (even with ``--dry-run``).
159 The release fixes that bug (nothing else changed).
165 *(Released Mon, 07 Mar 2011)*
167 Final 2.4.0 release. Just a few small fixes:
169 - Fixed RAPI node evacuate
170 - Fixed the kvm-ifup script
171 - Fixed internal error handling for special job cases
172 - Updated man page to specify the escaping feature for options
178 *(Released Mon, 28 Feb 2011)*
180 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
181 that this feature wiped the disks of the wrong instance, leading to loss
186 - Fixed title of query field containing instance name
187 - Expanded the glossary in the documentation
188 - Fixed one unittest (internal issue)
194 *(Released Mon, 21 Feb 2011)*
196 A number of bug fixes plus just a couple functionality changes.
198 On the user-visible side, the ``gnt-* list`` command output has changed
199 with respect to "special" field states. The current rc1 style of display
200 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
201 the default output mode special fields states are displayed as follows:
203 - Offline resource: ``*``
204 - Unavailable/not applicable: ``-``
205 - Data missing (RPC failure): ``?``
206 - Unknown field: ``??``
208 Another user-visible change is the addition of ``--force-join`` to
213 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
214 - Fixed regression in RAPI/instance reinstall where all parameters were
215 required (instead of optional)
216 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
217 - Fixed iallocator usage (offline nodes were not considered offline)
218 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
219 - Fixed hypervisor and OS parameter validation with respect to
221 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
223 - Fixed ``tools/listrunner`` with respect to agent-based usage
229 *(Released Fri, 4 Feb 2011)*
231 Many changes and fixes since the beta1 release. While there were some
232 internal changes, the code has been mostly stabilised for the RC
235 Note: the dumb allocator was removed in this release, as it was not kept
236 up-to-date with the IAllocator protocol changes. It is recommended to
237 use the ``hail`` command from the ganeti-htools package.
239 Note: the 2.4 and up versions of Ganeti are not compatible with the
240 0.2.x branch of ganeti-htools. You need to upgrade to
241 ganeti-htools-0.3.0 (or later).
243 Regressions fixed from 2.3
244 ~~~~~~~~~~~~~~~~~~~~~~~~~~
246 - Fixed the ``gnt-cluster verify-disks`` command
247 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
249 - Fixed disk adoption breakage
250 - Fixed wrong headers in instance listing for field aliases
255 - Fixed corner case in KVM handling of NICs
256 - Fixed many cases of wrong handling of non-vm_capable nodes
257 - Fixed a bug where a missing instance symlink was not possible to
258 recreate with any ``gnt-*`` command (now ``gnt-instance
259 activate-disks`` does it)
260 - Fixed the volume group name as reported by ``gnt-cluster
262 - Increased timeouts for the import-export code, hopefully leading to
263 fewer aborts due network or instance timeouts
264 - Fixed bug in ``gnt-node list-storage``
265 - Fixed bug where not all daemons were started on cluster
266 initialisation, but only at the first watcher run
267 - Fixed many bugs in the OOB implementation
268 - Fixed watcher behaviour in presence of instances with offline
270 - Fixed instance list output for instances running on the wrong node
271 - a few fixes to the cluster-merge tool, but it still cannot merge
272 multi-node groups (currently it is not recommended to use this tool)
278 - Improved network configuration for the KVM hypervisor
279 - Added e1000 as a supported NIC for Xen-HVM
280 - Improved the lvmstrap tool to also be able to use partitions, as
281 opposed to full disks
282 - Improved speed of disk wiping (the cluster parameter
283 ``prealloc_wipe_disks``, so that it has a low impact on the total time
284 of instance creations
285 - Added documentation for the OS parameters
286 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
287 hypervisor is not responding
288 - Added display of blacklisted and hidden OS information in
290 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
291 NIC and node parameters, which might create problems with currently
292 invalid (but undetected) configuration files, but prevents validation
293 failures when unrelated parameters are modified
294 - Changed cluster initialisation to wait for the master daemon to become
296 - Expanded the RAPI interface:
298 - Added config redistribution resource
299 - Added activation/deactivation of instance disks
300 - Added export of console information
302 - Implemented log file reopening on SIGHUP, which allows using
303 logrotate(8) for the Ganeti log files
304 - Added a basic OOB helper script as an example
310 *(Released Fri, 14 Jan 2011)*
315 - Fixed timezone issues when formatting timestamps
316 - Added support for node groups, available via ``gnt-group`` and other
318 - Added out-of-band framework and management, see :doc:`design
319 document <design-oob>`
320 - Removed support for roman numbers from ``gnt-node list`` and
321 ``gnt-instance list``.
322 - Allowed modification of master network interface via ``gnt-cluster
323 modify --master-netdev``
324 - Accept offline secondaries while shutting down instance disks
325 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
326 - Added support for multiple LVM volume groups
327 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
329 - Added commands to list available fields:
331 - ``gnt-node list-fields``
332 - ``gnt-group list-fields``
333 - ``gnt-instance list-fields``
335 - Updated documentation and man pages
340 - Moved ``rapi_users`` file into separate directory, now named
341 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
343 - Added new tool for running commands on many machines,
344 ``tools/ganeti-listrunner``
345 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
346 improving the ``gnt-instance console`` output
347 - Allowed customisation of disk index separator at ``configure`` time
348 - Export node group allocation policy to :doc:`iallocator <iallocator>`
349 - Added support for non-partitioned md disks in ``lvmstrap``
350 - Added script to gracefully power off KVM instances
351 - Split ``utils`` module into smaller parts
352 - Changed query operations to return more detailed information, e.g.
353 whether an information is unavailable due to an offline node. To use
354 this new functionality, the LUXI call ``Query`` must be used. Field
355 information is now stored by the master daemon and can be retrieved
356 using ``QueryFields``. Instances, nodes and groups can also be queried
357 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
358 exposed via RAPI). The following commands make use of this
359 infrastructure change:
362 - ``gnt-group list-fields``
364 - ``gnt-node list-fields``
365 - ``gnt-instance list``
366 - ``gnt-instance list-fields``
367 - ``gnt-debug locks``
372 - New RAPI resources (see :doc:`rapi`):
376 - ``/2/groups/[group_name]``
377 - ``/2/groups/[group_name]/assign-nodes``
378 - ``/2/groups/[group_name]/modify``
379 - ``/2/groups/[group_name]/rename``
380 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
384 - Implemented ``no_install`` for instance creation
385 - Implemented OS parameters for instance reinstallation, allowing
386 use of special settings on reinstallation (e.g. for preserving data)
391 - Added IPv6 support in import/export
392 - Pause DRBD synchronization while wiping disks on instance creation
393 - Updated unittests and QA scripts
394 - Improved network parameters passed to KVM
395 - Converted man pages from docbook to reStructuredText
401 *(Released Mon, 20 Dec 2010)*
403 Released version 2.3.1~rc1 without any changes.
409 *(Released Wed, 1 Dec 2010)*
411 - impexpd: Disable OpenSSL compression in socat if possible (backport
412 from master, commit e90739d625b, see :doc:`installation guide
413 <install-quick>` for details)
414 - Changed unittest coverage report to exclude test scripts
415 - Added script to check version format
421 *(Released Wed, 1 Dec 2010)*
423 Released version 2.3.0~rc1 without any changes.
429 *(Released Fri, 19 Nov 2010)*
431 A number of bugfixes and documentation updates:
433 - Update ganeti-os-interface documentation
434 - Fixed a bug related to duplicate MACs or similar items which should be
436 - Fix breakage in OS state modify
437 - Reinstall instance: disallow offline secondaries (fixes bug related to
438 OS changing but reinstall failing)
439 - plus all the other fixes between 2.2.1 and 2.2.2
445 *(Released Tue, 2 Nov 2010)*
447 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
448 - Fixed master failover race with watcher
449 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
451 - Accept previously stopped instance for export with instance removal
452 - Simplify and extend the environment variables for instance OS scripts
453 - Added new node flags, ``master_capable`` and ``vm_capable``
454 - Added optional instance disk wiping prior during allocation. This is a
455 cluster-wide option and can be set/modified using
456 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
457 - Added IPv6 support, see :doc:`design document <design-2.3>` and
459 - Added a new watcher option (``--ignore-pause``)
460 - Added option to ignore offline node on instance start/stop
461 (``--ignore-offline``)
462 - Allow overriding OS parameters with ``gnt-instance reinstall``
463 - Added ability to change node's secondary IP address using ``gnt-node
465 - Implemented privilege separation for all daemons except
466 ``ganeti-noded``, see ``configure`` options
467 - Complain if an instance's disk is marked faulty in ``gnt-cluster
469 - Implemented job priorities (see ``ganeti(7)`` manpage)
470 - Ignore failures while shutting down instances during failover from
472 - Exit daemon's bootstrap process only once daemon is ready
473 - Export more information via ``LUInstanceQuery``/remote API
474 - Improved documentation, QA and unittests
475 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
476 restart if the file was created or changed
477 - Added LUXI protocol version sent with each request and response,
478 allowing detection of server/client mismatches
479 - Moved the Python scripts among gnt-* and ganeti-* into modules
480 - Moved all code related to setting up SSH to an external script,
482 - Infrastructure changes for node group support in future versions
488 *(Released Fri, 19 Nov 2010)*
490 A few small bugs fixed, and some improvements to the build system:
492 - Fix documentation regarding conversion to drbd
493 - Fix validation of parameters in cluster modify (``gnt-cluster modify
495 - Fix error handling in node modify with multiple changes
496 - Allow remote imports without checked names
502 *(Released Tue, 19 Oct 2010)*
504 - Disable SSL session ID cache in RPC client
510 *(Released Thu, 14 Oct 2010)*
512 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
513 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
515 - Fix problems with interaction between Curl and Python's HTTP server,
516 resulting in increased speed in many RPC calls
517 - Improve our release script to prevent breakage with older aclocal and
524 *(Released Thu, 7 Oct 2010)*
526 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
527 value retrieved from master
528 - Added support for blacklisted or hidden OS definitions
529 - Added simple lock monitor (accessible via (``gnt-debug locks``)
530 - Added support for -mem-path in KVM hypervisor abstraction layer
531 - Allow overriding instance parameters in tool for inter-cluster
532 instance moves (``tools/move-instance``)
533 - Improved opcode summaries (e.g. in ``gnt-job list``)
534 - Improve consistency of OS listing by sorting it
535 - Documentation updates
541 *(Released Fri, 8 Oct 2010)*
543 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
549 *(Released Mon, 4 Oct 2010)*
551 - Fixed regression in ``gnt-instance rename``
557 *(Released Wed, 22 Sep 2010)*
559 - Fixed OS_VARIANT variable for OS scripts
560 - Fixed cluster tag operations via RAPI
561 - Made ``setup-ssh`` exit with non-zero code if an error occurred
562 - Disabled RAPI CA checks in watcher
568 *(Released Mon, 23 Aug 2010)*
570 - Support DRBD versions of the format "a.b.c.d"
572 - Re-introduce support for usage from multiple threads in RAPI client
573 - Instance renames and modify via RAPI
574 - Work around race condition between processing and archival in job
576 - Mark opcodes following failed one as failed, too
577 - Job field ``lock_status`` was removed due to difficulties making it
578 work with the changed job queue in Ganeti 2.2; a better way to monitor
579 locks is expected for a later 2.2.x release
580 - Fixed dry-run behaviour with many commands
581 - Support ``ssh-agent`` again when adding nodes
582 - Many additional bugfixes
588 *(Released Fri, 30 Jul 2010)*
590 Important change: the internal RPC mechanism between Ganeti nodes has
591 changed from using a home-grown http library (based on the Python base
592 libraries) to use the PycURL library. This requires that PycURL is
593 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
594 against GnuTLS by default. cURL's support for GnuTLS had known issues
595 before cURL 7.21.0 and we recommend using the latest cURL release or
596 linking against OpenSSL. Most other distributions already link PycURL
597 and cURL against OpenSSL. The command::
599 python -c 'import pycurl; print pycurl.version'
601 can be used to determine the libraries PycURL and cURL are linked
604 Other significant changes:
606 - Rewrote much of the internals of the job queue, in order to achieve
607 better parallelism; this decouples job query operations from the job
608 processing, and it should allow much nicer behaviour of the master
609 daemon under load, and it also has uncovered some long-standing bugs
610 related to the job serialisation (now fixed)
611 - Added a default iallocator setting to the cluster parameters,
612 eliminating the need to always pass nodes or an iallocator for
613 operations that require selection of new node(s)
614 - Added experimental support for the LXC virtualization method
615 - Added support for OS parameters, which allows the installation of
616 instances to pass parameter to OS scripts in order to customise the
618 - Added a hypervisor parameter controlling the migration type (live or
619 non-live), since hypervisors have various levels of reliability; this
620 has renamed the 'live' parameter to 'mode'
621 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
622 logical volumes, meaning that cluster verify will ignore them and not
623 flag their presence as errors
624 - The watcher will now reset the error count for failed instances after
625 8 hours, thus allowing self-healing if the problem that caused the
626 instances to be down/fail to start has cleared in the meantime
627 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
628 check for, and warn, if the drbd module parameter ``usermode_helper``
629 is not consistent with the cluster-wide setting; this is needed to
630 make diagnose easier of failed drbd creations
631 - Started adding base IPv6 support, but this is not yet
632 enabled/available for use
633 - Rename operations (cluster, instance) will now return the new name,
634 which is especially useful if a short name was passed in
635 - Added support for instance migration in RAPI
636 - Added a tool to pre-configure nodes for the SSH setup, before joining
637 them to the cluster; this will allow in the future a simplified model
638 for node joining (but not yet fully enabled in 2.2); this needs the
639 paramiko python library
640 - Fixed handling of name-resolving errors
641 - Fixed consistency of job results on the error path
642 - Fixed master-failover race condition when executed multiple times in
644 - Fixed many bugs related to the job queue (mostly introduced during the
645 2.2 development cycle, so not all are impacting 2.1)
646 - Fixed instance migration with missing disk symlinks
647 - Fixed handling of unknown jobs in ``gnt-job archive``
648 - And many other small fixes/improvements
652 - Enhanced both the unittest and the QA coverage
653 - Switched the opcode validation to a generic model, and extended the
654 validation to all opcode parameters
655 - Changed more parts of the code that write shell scripts to use the
657 - Switched the master daemon to use the asyncore library for the Luxi
664 *(Released Thu, 17 Jun 2010)*
666 - Added tool (``move-instance``) and infrastructure to move instances
667 between separate clusters (see :doc:`separate documentation
668 <move-instance>` and :doc:`design document <design-2.2>`)
669 - Added per-request RPC timeout
670 - RAPI now requires a Content-Type header for requests with a body (e.g.
671 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
672 :rfc:`2616` (HTTP/1.1), section 7.2.1)
673 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
675 - Implemented initial support for running Ganeti daemons as separate
676 users, see configure-time flags ``--with-user-prefix`` and
677 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
679 - Instances can be removed after export (``gnt-backup export
681 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
682 key (instead of 1024 bit)
683 - Added new cluster configuration file for cluster domain secret
684 - Import/export now use SSL instead of SSH
685 - Added support for showing estimated time when exporting an instance,
686 see the ``ganeti-os-interface(7)`` manpage and look for
693 *(Released Tue, 16 Nov 2010)*
695 Some more bugfixes. Unless critical bugs occur, this will be the last
698 - Fix case of MAC special-values
699 - Fix mac checker regex
700 - backend: Fix typo causing "out of range" error
701 - Add missing --units in gnt-instance list man page
707 *(Released Tue, 24 Aug 2010)*
710 - Don't ignore secondary node silently on non-mirrored disk templates
712 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
713 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
714 - Properly document the usb_mouse parameter
715 - Fix path in ganeti-rapi(8) (issue 116)
716 - Adjust error message when the ganeti user's .ssh directory is
718 - Add same-node-check when changing the disk template to drbd
724 *(Released Fri, 16 Jul 2010)*
727 - Add an option to only select some reboot types during qa/burnin.
728 (on some hypervisors consequent reboots are not supported)
729 - Fix infrequent race condition in master failover. Sometimes the old
730 master ip address would be still detected as up for a short time
731 after it was removed, causing failover to fail.
732 - Decrease mlockall warnings when the ctypes module is missing. On
733 Python 2.4 we support running even if no ctypes module is installed,
734 but we were too verbose about this issue.
735 - Fix building on old distributions, on which man doesn't have a
737 - Fix RAPI not to ignore the MAC address on instance creation
738 - Implement the old instance creation format in the RAPI client.
744 *(Released Thu, 01 Jul 2010)*
746 A small bugfix release:
747 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
748 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
749 a non-existing option
750 - Fix instance create: the --force-variant option was ignored
751 - Improve pylint 0.21 compatibility and warnings with Python 2.6
752 - Fix modify node storage with non-FQDN arguments
753 - Fix RAPI client to authenticate under Python 2.6 when used
754 for more than 5 requests needing authentication
755 - Fix gnt-instance modify -t (storage) giving a wrong error message
756 when converting a non-shutdown drbd instance to plain
762 *(Released Fri, 18 Jun 2010)*
764 A small bugfix release:
766 - Fix live migration of KVM instances started with older Ganeti
767 versions which had fewer hypervisor parameters
768 - Fix gnt-instance grow-disk on down instances
769 - Fix an error-reporting bug during instance migration
770 - Better checking of the ``--net`` and ``--disk`` values, to avoid
771 silently ignoring broken ones
772 - Fix an RPC error reporting bug affecting, for example, RAPI client
774 - Fix bug triggered by different API version os-es on different nodes
775 - Fix a bug in instance startup with custom hvparams: OS level
776 parameters would fail to be applied.
777 - Fix the RAPI client under Python 2.6 (but more work is needed to
778 make it work completely well with OpenSSL)
779 - Fix handling of errors when resolving names from DNS
785 *(Released Thu, 3 Jun 2010)*
787 A medium sized development cycle. Some new features, and some
788 fixes/small improvements/cleanups.
793 The node deamon now tries to mlock itself into memory, unless the
794 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
795 its logs, and falls back to console logging. This allows emergency
796 features such as ``gnt-node powercycle`` to work even in the event of a
797 broken node disk (tested offlining the disk hosting the node's
798 filesystem and dropping its memory caches; don't try this at home)
800 KVM: add vhost-net acceleration support. It can be tested with a new
801 enough version of the kernel and of qemu-kvm.
803 KVM: Add instance chrooting feature. If you use privilege dropping for
804 your VMs you can also now force them to chroot to an empty directory,
805 before starting the emulated guest.
807 KVM: Add maximum migration bandwith and maximum downtime tweaking
808 support (requires a new-enough version of qemu-kvm).
810 Cluster verify will now warn if the master node doesn't have the master
813 Add a new (incompatible) instance creation request format to RAPI which
814 supports all parameters (previously only a subset was supported, and it
815 wasn't possible to extend the old format to accomodate all the new
816 features. The old format is still supported, and a client can check for
817 this feature, before using it, by checking for its presence in the
818 ``features`` RAPI resource.
820 Now with ancient latin support. Try it passing the ``--roman`` option to
821 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
822 (requires the python-roman module to be installed, in order to work).
827 As usual many internal code refactorings, documentation updates, and
830 - Lots of improvements and cleanups to the experimental Remote API
831 (RAPI) client library.
832 - A new unit test suite for the core daemon libraries.
833 - A fix to creating missing directories makes sure the umask is not
834 applied anymore. This enforces the same directory permissions
836 - Better handling terminating daemons with ctrl+c (used when running
837 them in debugging mode).
838 - Fix a race condition in live migrating a KVM instance, when stat()
839 on the old proc status file returned EINVAL, which is an unexpected
841 - Fixed manpage checking with newer man and utf-8 charachters. But now
842 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
848 *(Released Fri, 7 May 2010)*
850 Fix a bug which prevented untagged KVM instances from starting.
856 *(Released Fri, 7 May 2010)*
858 Another release with a long development cycle, during which many
859 different features were added.
864 The KVM hypervisor now can run the individual instances as non-root, to
865 reduce the impact of a VM being hijacked due to bugs in the
866 hypervisor. It is possible to run all instances as a single (non-root)
867 user, to manually specify a user for each instance, or to dynamically
868 allocate a user out of a cluster-wide pool to each instance, with the
869 guarantee that no two instances will run under the same user ID on any
872 An experimental RAPI client library, that can be used standalone
873 (without the other Ganeti libraries), is provided in the source tree as
874 ``lib/rapi/client.py``. Note this client might change its interface in
875 the future, as we iterate on its capabilities.
877 A new command, ``gnt-cluster renew-crypto`` has been added to easily
878 replace the cluster's certificates and crypto keys. This might help in
879 case they have been compromised, or have simply expired.
881 A new disk option for instance creation has been added that allows one
882 to "adopt" currently existing logical volumes, with data
883 preservation. This should allow easier migration to Ganeti from
884 unmanaged (or managed via other software) instances.
886 Another disk improvement is the possibility to convert between redundant
887 (DRBD) and plain (LVM) disk configuration for an instance. This should
888 allow better scalability (starting with one node and growing the
889 cluster, or shrinking a two-node cluster to one node).
891 A new feature that could help with automated node failovers has been
892 implemented: if a node sees itself as offline (by querying the master
893 candidates), it will try to shutdown (hard) all instances and any active
894 DRBD devices. This reduces the risk of duplicate instances if an
895 external script automatically failovers the instances on such nodes. To
896 enable this, the cluster parameter ``maintain_node_health`` should be
897 enabled; in the future this option (per the name) will enable other
898 automatic maintenance features.
900 Instance export/import now will reuse the original instance
901 specifications for all parameters; that means exporting an instance,
902 deleting it and the importing it back should give an almost identical
903 instance. Note that the default import behaviour has changed from
904 before, where it created only one NIC; now it recreates the original
907 Cluster verify has added a few new checks: SSL certificates validity,
908 /etc/hosts consistency across the cluster, etc.
913 As usual, many internal changes were done, documentation fixes,
916 - Fixed cluster initialization with disabled cluster storage (regression
918 - File-based storage supports growing the disks
919 - Fixed behaviour of node role changes
920 - Fixed cluster verify for some corner cases, plus a general rewrite of
921 cluster verify to allow future extension with more checks
922 - Fixed log spamming by watcher and node daemon (regression introduced
924 - Fixed possible validation issues when changing the list of enabled
926 - Fixed cleanup of /etc/hosts during node removal
927 - Fixed RAPI response for invalid methods
928 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
929 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
930 from ide disks, etc.)
931 - Allow OS changes without re-installation (to record a changed OS
932 outside of Ganeti, or to allow OS renames)
933 - Allow instance creation without OS installation (useful for example if
934 the OS will be installed manually, or restored from a backup not in
936 - Implemented option to make cluster ``copyfile`` use the replication
938 - Added list of enabled hypervisors to ssconf (possibly useful for
940 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
942 - A partial form of node re-IP is possible via node readd, which now
943 allows changed node primary IP
944 - Command line utilities now show an informational message if the job is
946 - The logs of the master daemon now show the PID/UID/GID of the
953 *(Released Fri, 12 Mar 2010)*
955 During the 2.1.0 long release candidate cycle, a lot of improvements and
956 changes have accumulated with were released later as 2.1.1.
961 The node evacuate command (``gnt-node evacuate``) was significantly
962 rewritten, and as such the IAllocator protocol was changed - a new
963 request type has been added. This unfortunate change during a stable
964 series is designed to improve performance of node evacuations; on
965 clusters with more than about five nodes and which are well-balanced,
966 evacuation should proceed in parallel for all instances of the node
967 being evacuated. As such, any existing IAllocator scripts need to be
968 updated, otherwise the above command will fail due to the unknown
969 request. The provided "dumb" allocator has not been updated; but the
970 ganeti-htools package supports the new protocol since version 0.2.4.
972 Another important change is increased validation of node and instance
973 names. This might create problems in special cases, if invalid host
974 names are being used.
976 Also, a new layer of hypervisor parameters has been added, that sits at
977 OS level between the cluster defaults and the instance ones. This allows
978 customisation of virtualization parameters depending on the installed
979 OS. For example instances with OS 'X' may have a different KVM kernel
980 (or any other parameter) than the cluster defaults. This is intended to
981 help managing a multiple OSes on the same cluster, without manual
982 modification of each instance's parameters.
984 A tool for merging clusters, ``cluster-merge``, has been added in the
990 - Improved the int/float conversions that should make the code more
991 robust in face of errors from the node daemons
992 - Fixed the remove node code in case of internal configuration errors
993 - Fixed the node daemon behaviour in face of inconsistent queue
994 directory (e.g. read-only file-system where we can't open the files
996 - Fixed the behaviour of gnt-node modify for master candidate demotion;
997 now it either aborts cleanly or, if given the new "auto_promote"
998 parameter, will automatically promote other nodes as needed
999 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1000 completely prevent Ganeti from working
1001 - Fixed bug for instance export when not all disks were successfully
1003 - Fixed behaviour of node add when the new node is slow in starting up
1005 - Fixed handling of signals in the LUXI client, which should improve
1006 behaviour of command-line scripts
1007 - Added checks for invalid node/instance names in the configuration (now
1008 flagged during cluster verify)
1009 - Fixed watcher behaviour for disk activation errors
1010 - Fixed two potentially endless loops in http library, which led to the
1011 RAPI daemon hanging and consuming 100% CPU in some cases
1012 - Fixed bug in RAPI daemon related to hashed passwords
1013 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1015 - Enhanced compatibility with non-Debian OSes, but not using absolute
1016 path in some commands and allowing customisation of the ssh
1017 configuration directory
1018 - Fixed possible future issue with new Python versions by abiding to the
1019 proper use of ``__slots__`` attribute on classes
1020 - Added checks that should prevent directory traversal attacks
1021 - Many documentation fixes based on feedback from users
1026 - Added an "early_release" more for instance replace disks and node
1027 evacuate, where we release locks earlier and thus allow higher
1028 parallelism within the cluster
1029 - Added watcher hooks, intended to allow the watcher to restart other
1030 daemons (e.g. from the ganeti-nbma project), but they can be used of
1031 course for any other purpose
1032 - Added a compile-time disable for DRBD barriers, to increase
1033 performance if the administrator trusts the power supply or the
1034 storage system to not lose writes
1035 - Added the option of using syslog for logging instead of, or in
1036 addition to, Ganeti's own log files
1037 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1038 can indeed boot from a paravirtual NIC
1039 - Added a generic debug level for many operations; while this is not
1040 used widely yet, it allows one to pass the debug value all the way to
1042 - Enhanced the hooks environment for instance moves (failovers,
1043 migrations) where the primary/secondary nodes changed during the
1044 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1045 - Enhanced data validations for many user-supplied values; one important
1046 item is the restrictions imposed on instance and node names, which
1047 might reject some (invalid) host names
1048 - Add a configure-time option to disable file-based storage, if it's not
1049 needed; this allows greater security separation between the master
1050 node and the other nodes from the point of view of the inter-node RPC
1052 - Added user notification in interactive tools if job is waiting in the
1053 job queue or trying to acquire locks
1054 - Added log messages when a job is waiting for locks
1055 - Added filtering by node tags in instance operations which admit
1056 multiple instances (start, stop, reboot, reinstall)
1057 - Added a new tool for cluster mergers, ``cluster-merge``
1058 - Parameters from command line which are of the form ``a=b,c=d`` can now
1059 use backslash escapes to pass in values which contain commas,
1060 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1062 - For KVM, the instance name is the first parameter passed to KVM, so
1063 that it's more visible in the process list
1069 *(Released Tue, 2 Mar 2010)*
1071 Ganeti 2.1 brings many improvements with it. Major changes:
1073 - Added infrastructure to ease automated disk repairs
1074 - Added new daemon to export configuration data in a cheaper way than
1075 using the remote API
1076 - Instance NICs can now be routed instead of being associated with a
1078 - Improved job locking logic to reduce impact of jobs acquiring multiple
1079 locks waiting for other long-running jobs
1081 In-depth implementation details can be found in the Ganeti 2.1 design
1087 - Added chroot hypervisor
1088 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1090 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1091 ``bootloader_path`` and ``bootloader_args``)
1092 - Added the ``use_localtime`` option for the xen-hvm and kvm
1093 hypervisors, and the default value for this has changed to false (in
1094 2.0 xen-hvm always enabled it)
1095 - Added luxi call to submit multiple jobs in one go
1096 - Added cluster initialization option to not modify ``/etc/hosts``
1098 - Added network interface parameters
1099 - Added dry run mode to some LUs
1100 - Added RAPI resources:
1102 - ``/2/instances/[instance_name]/info``
1103 - ``/2/instances/[instance_name]/replace-disks``
1104 - ``/2/nodes/[node_name]/evacuate``
1105 - ``/2/nodes/[node_name]/migrate``
1106 - ``/2/nodes/[node_name]/role``
1107 - ``/2/nodes/[node_name]/storage``
1108 - ``/2/nodes/[node_name]/storage/modify``
1109 - ``/2/nodes/[node_name]/storage/repair``
1111 - Added OpCodes to evacuate or migrate all instances on a node
1112 - Added new command to list storage elements on nodes (``gnt-node
1113 list-storage``) and modify them (``gnt-node modify-storage``)
1114 - Added new ssconf files with master candidate IP address
1115 (``ssconf_master_candidates_ips``), node primary IP address
1116 (``ssconf_node_primary_ips``) and node secondary IP address
1117 (``ssconf_node_secondary_ips``)
1118 - Added ``ganeti-confd`` and a client library to query the Ganeti
1119 configuration via UDP
1120 - Added ability to run hooks after cluster initialization and before
1122 - Added automatic mode for disk replace (``gnt-instance replace-disks
1124 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1125 after catastrophic data-loss
1126 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1128 - Added ``gnt-instance move`` command to move instances
1129 - Added ``gnt-cluster watcher`` command to control watcher
1130 - Added ``gnt-node powercycle`` command to powercycle nodes
1131 - Added new job status field ``lock_status``
1132 - Added parseable error codes to cluster verification (``gnt-cluster
1133 verify --error-codes``) and made output less verbose (use
1134 ``--verbose`` to restore previous behaviour)
1135 - Added UUIDs to the main config entities (cluster, nodes, instances)
1136 - Added support for OS variants
1137 - Added support for hashed passwords in the Ganeti remote API users file
1139 - Added option to specify maximum timeout on instance shutdown
1140 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1141 - Added new helper script to start and stop Ganeti daemons
1142 (``daemon-util``), with the intent to reduce the work necessary to
1143 adjust Ganeti for non-Debian distributions and to start/stop daemons
1145 - Added more unittests
1146 - Fixed critical bug in ganeti-masterd startup
1147 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1148 now customisable at the cluster level for both the KVM and Xen
1149 hypervisors using the new ``migration_port`` parameter
1150 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1151 reinstalling an instance
1152 - Allowed ``@`` in tag names
1153 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1154 - Many documentation updates
1155 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1156 - ``gnt-instance reinstall`` can now reinstall multiple instances
1157 - Updated many command line parameters
1158 - Introduced new OS API version 15
1159 - No longer support a default hypervisor
1160 - Treat virtual LVs as inexistent
1161 - Improved job locking logic to reduce lock contention
1162 - Match instance and node names case insensitively
1163 - Reimplemented bash completion script to be more complete
1170 *(Released Thu, 4 Feb 2010)*
1172 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1173 - Fix a string formatting bug
1174 - Improve safety of the code in some error paths
1175 - Improve data validation in the master of values returned from nodes
1181 *(Released Thu, 17 Dec 2009)*
1183 - Fix security issue due to missing validation of iallocator names; this
1184 allows local and remote execution of arbitrary executables
1185 - Fix failure of gnt-node list during instance removal
1186 - Ship the RAPI documentation in the archive
1192 *(Released Wed, 30 Sep 2009)*
1194 - Fixed many wrong messages
1195 - Fixed a few bugs related to the locking library
1196 - Fixed MAC checking at instance creation time
1197 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1198 - Fixed a few issues related to signal handling in both daemons and
1200 - Fixed the example startup script provided
1201 - Fixed insserv dependencies in the example startup script (patch from
1203 - Fixed handling of drained nodes in the iallocator framework
1204 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1206 - Fixed error related to invalid job IDs in job polling
1207 - Fixed job/opcode persistence on unclean master shutdown
1208 - Fixed handling of partial job processing after unclean master
1210 - Fixed error reporting from LUs, previously all errors were converted
1211 into execution errors
1212 - Fixed error reporting from burnin
1213 - Decreased significantly the memory usage of the job queue
1214 - Optimised slightly multi-job submission
1215 - Optimised slightly opcode loading
1216 - Backported the multi-job submit framework from the development
1217 branch; multi-instance start and stop should be faster
1218 - Added script to clean archived jobs after 21 days; this will reduce
1219 the size of the queue directory
1220 - Added some extra checks in disk size tracking
1221 - Added an example ethers hook script
1222 - Added a cluster parameter that prevents Ganeti from modifying of
1224 - Added more node information to RAPI responses
1225 - Added a ``gnt-job watch`` command that allows following the ouput of a
1227 - Added a bind-address option to ganeti-rapi
1228 - Added more checks to the configuration verify
1229 - Enhanced the burnin script such that some operations can be retried
1231 - Converted instance reinstall to multi-instance model
1237 *(Released Fri, 7 Aug 2009)*
1239 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1240 to allow using the pre-2.0.2 behaviour in activation, if any existing
1241 instances have mismatched disk sizes in the configuration
1242 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1243 any configuration mismatches for disk sizes
1244 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1245 failover to work on two-node clusters
1246 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1248 - Fixed detection of OS script errors in ``gnt-backup export``
1249 - Fixed exit code of ``gnt-backup export``
1255 *(Released Fri, 17 Jul 2009)*
1257 - Added experimental support for stripped logical volumes; this should
1258 enhance performance but comes with a higher complexity in the block
1259 device handling; stripping is only enabled when passing
1260 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1261 affected even in the non-stripped mode
1262 - Improved resiliency against transient failures at the end of DRBD
1263 resyncs, and in general of DRBD resync checks
1264 - Fixed a couple of issues with exports and snapshot errors
1265 - Fixed a couple of issues in instance listing
1266 - Added display of the disk size in ``gnt-instance info``
1267 - Fixed checking for valid OSes in instance creation
1268 - Fixed handling of the "vcpus" parameter in instance listing and in
1269 general of invalid parameters
1270 - Fixed http server library, and thus RAPI, to handle invalid
1271 username/password combinations correctly; this means that now they
1272 report unauthorized for queries too, not only for modifications,
1273 allowing earlier detect of configuration problems
1274 - Added a new "role" node list field, equivalent to the master/master
1275 candidate/drained/offline flags combinations
1276 - Fixed cluster modify and changes of candidate pool size
1277 - Fixed cluster verify error messages for wrong files on regular nodes
1278 - Fixed a couple of issues with node demotion from master candidate role
1279 - Fixed node readd issues
1280 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1281 - Added a new ``--no-voting`` option for masterfailover to fix failover
1282 on two-nodes clusters when the former master node is unreachable
1283 - Added instance reinstall over RAPI
1289 *(Released Tue, 16 Jun 2009)*
1291 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1292 allow re-adding the start in single-user option (regression from 1.2)
1293 - the watcher writes the instance status to a file, to allow monitoring
1294 to report the instance status (from the master) based on cached
1295 results of the watcher's queries; while this can get stale if the
1296 watcher is being locked due to other work on the cluster, this is
1297 still an improvement
1298 - the watcher now also restarts the node daemon and the rapi daemon if
1300 - fixed the watcher to handle full and drained queue cases
1301 - hooks export more instance data in the environment, which helps if
1302 hook scripts need to take action based on the instance's properties
1303 (no longer need to query back into ganeti)
1304 - instance failovers when the instance is stopped do not check for free
1305 RAM, so that failing over a stopped instance is possible in low memory
1307 - rapi uses queries for tags instead of jobs (for less job traffic), and
1308 for cluster tags it won't talk to masterd at all but read them from
1310 - a couple of error handling fixes in RAPI
1311 - drbd handling: improved the error handling of inconsistent disks after
1312 resync to reduce the frequency of "there are some degraded disks for
1313 this instance" messages
1314 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1315 error handling path called a wrong function name)
1321 *(Released Wed, 27 May 2009)*
1323 - no changes from rc5
1326 Version 2.0 release candidate 5
1327 -------------------------------
1329 *(Released Wed, 20 May 2009)*
1331 - fix a couple of bugs (validation, argument checks)
1332 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1333 - some small improvements to RAPI and IAllocator
1334 - make watcher automatically start the master daemon if down
1337 Version 2.0 release candidate 4
1338 -------------------------------
1340 *(Released Mon, 27 Apr 2009)*
1342 - change the OS list to not require locks; this helps with big clusters
1343 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1344 volume group is broken
1345 - ``gnt-instance info``, without any arguments, doesn't run for all
1346 instances anymore; either pass ``--all`` or pass the desired
1347 instances; this helps against mistakes on big clusters where listing
1348 the information for all instances takes a long time
1349 - miscellaneous doc and man pages fixes
1352 Version 2.0 release candidate 3
1353 -------------------------------
1355 *(Released Wed, 8 Apr 2009)*
1357 - Change the internal locking model of some ``gnt-node`` commands, in
1358 order to reduce contention (and blocking of master daemon) when
1359 batching many creation/reinstall jobs
1360 - Fixes to Xen soft reboot
1361 - No longer build documentation at build time, instead distribute it in
1362 the archive, in order to reduce the need for the whole docbook/rst
1366 Version 2.0 release candidate 2
1367 -------------------------------
1369 *(Released Fri, 27 Mar 2009)*
1371 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1372 - Fix watcher startup sequence, improves the behaviour of busy clusters
1373 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1374 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1375 bind address and other places
1376 - Some documentation fixes and updates
1379 Version 2.0 release candidate 1
1380 -------------------------------
1382 *(Released Mon, 2 Mar 2009)*
1384 - More documentation updates, now all docs should be more-or-less
1386 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1388 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1390 - Fix an issue related to $libdir/run/ganeti and cluster creation
1396 *(Released Thu, 19 Feb 2009)*
1398 - Xen PVM and KVM have switched the default value for the instance root
1399 disk to the first partition on the first drive, instead of the whole
1400 drive; this means that the OS installation scripts must be changed
1402 - Man pages have been updated
1403 - RAPI has been switched by default to HTTPS, and the exported functions
1404 should all work correctly
1405 - RAPI v1 has been removed
1406 - Many improvements to the KVM hypervisor
1407 - Block device errors are now better reported
1408 - Many other bugfixes and small improvements
1414 *(Released Mon, 26 Jan 2009)*
1416 - Version 2 is a general rewrite of the code and therefore the
1417 differences are too many to list, see the design document for 2.0 in
1418 the ``doc/`` subdirectory for more details
1419 - In this beta version there is not yet a migration path from 1.2 (there
1420 will be one in the final 2.0 release)
1421 - A few significant changes are:
1423 - all commands are executed by a daemon (``ganeti-masterd``) and the
1424 various ``gnt-*`` commands are just front-ends to it
1425 - all the commands are entered into, and executed from a job queue,
1426 see the ``gnt-job(8)`` manpage
1427 - the RAPI daemon supports read-write operations, secured by basic
1428 HTTP authentication on top of HTTPS
1429 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1430 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1431 to migrate to DRBD 8 first while still running Ganeti 1.2)
1432 - DRBD devices are using statically allocated minor numbers, which
1433 will be assigned to existing instances during the migration process
1434 - there is support for both Xen PVM and Xen HVM instances running on
1436 - KVM virtualization is supported too
1437 - file-based storage has been implemented, which means that it is
1438 possible to run the cluster without LVM and DRBD storage, for
1439 example using a shared filesystem exported from shared storage (and
1440 still have live migration)
1446 *(Released Tue, 13 Jan 2009)*
1448 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1449 - Reuse the old instance mac address by default on instance import, if
1450 the instance name is the same.
1451 - Handle situations in which the node info rpc returns incomplete
1453 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1454 - Improved version of batcher:
1456 - state file support
1457 - instance mac address support
1458 - support for HVM clusters/instances
1460 - Add an option to show the number of cpu sockets and nodes in
1462 - Support OSes that handle more than one version of the OS api (but do
1463 not change the current API in any other way)
1464 - Fix ``gnt-node migrate``
1465 - ``gnt-debug`` man page
1466 - Fixes various more typos and small issues
1467 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1473 *(Released Wed, 24 Sep 2008)*
1475 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1476 type of disk exported to fully virtualized instances.
1477 - provide access to the serial console of HVM instances
1478 - instance auto_balance flag, set by default. If turned off it will
1479 avoid warnings on cluster verify if there is not enough memory to fail
1480 over an instance. in the future it will prevent automatically failing
1481 it over when we will support that.
1482 - batcher tool for instance creation, see ``tools/README.batcher``
1483 - ``gnt-instance reinstall --select-os`` to interactively select a new
1484 operating system when reinstalling an instance.
1485 - when changing the memory amount on instance modify a check has been
1486 added that the instance will be able to start. also warnings are
1487 emitted if the instance will not be able to fail over, if auto_balance
1489 - documentation fixes
1490 - sync fields between ``gnt-instance list/modify/add/import``
1491 - fix a race condition in drbd when the sync speed was set after giving
1492 the device a remote peer.
1498 *(Released Tue, 22 Jul 2008)*
1500 - note: the allowed size and number of tags per object were reduced
1501 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1502 - fixed twisted 8.x compatibility
1503 - fixed ``gnt-instance replace-disks`` with iallocator
1504 - add TCP keepalives on twisted connections to detect restarted nodes
1505 - disk increase support, see ``gnt-instance grow-disk``
1506 - implement bulk node/instance query for RAPI
1507 - add tags in node/instance listing (optional)
1508 - experimental migration (and live migration) support, read the man page
1509 for ``gnt-instance migrate``
1510 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1511 has some small improvements in handling its state file
1517 *(Released Fri, 13 Jun 2008)*
1519 - Experimental readonly, REST-based remote API implementation;
1520 automatically started on master node, TCP port 5080, if enabled by
1521 ``--enable-rapi`` parameter to configure script.
1522 - Instance allocator support. Add and import instance accept a
1523 ``--iallocator`` parameter, and call that instance allocator to decide
1524 which node to use for the instance. The iallocator document describes
1525 what's expected from an allocator script.
1526 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1527 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1528 node is lost there is still enough memory to fail over the instances
1530 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1531 ``gnt-cluster verify``, to check for site-specific compliance. All the
1532 hooks will run, and their output, if any, will be displayed. Any
1533 failing hook will make the verification return an error value.
1534 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1535 primary and secondary interfaces
1536 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1537 that is still declared as part of the cluster and has failed.
1538 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1539 specifying output fields, that just adds the chosen fields to the
1541 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1542 export from the filesystem.
1543 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1544 have been added. Using them you can enable/disable acpi and pae
1545 support, and specify a path for a cd image to be exported to the
1546 instance. These parameters as the name suggest only work on HVM
1548 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1549 PAE support will be set to the previously hardcoded values, but the
1550 (previously hardcoded) path to the CDROM ISO image will be unset and
1551 if required, needs to be set manually with ``gnt-instance modify``
1553 - The address to which an instance's VNC console is bound is now
1554 selectable per-instance, rather than being cluster wide. Of course
1555 this only applies to instances controlled via VNC, so currently just
1556 applies to HVM clusters.
1562 *(Released Mon, 18 Feb 2008)*
1564 - more tweaks to the disk activation code (especially helpful for DRBD)
1565 - change the default ``gnt-instance list`` output format, now there is
1566 one combined status field (see the manpage for the exact values this
1568 - some more fixes for the mac export to hooks change
1569 - make Ganeti not break with DRBD 8.2.x (which changed the version
1570 format in ``/proc/drbd``) (issue 24)
1571 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1572 template, allowing migration from DRBD0.7+MD to DRBD8
1578 *(Released Wed, 30 Jan 2008)*
1580 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1582 - add command aliases infrastructure and a few aliases
1583 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1584 man pages and the ``--help`` option of ``gnt-node
1585 list``/``gnt-instance list``
1586 - fix ``gnt-backup list`` with down nodes (issue 21)
1587 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1588 - fix the dist archive and add a check for including svn/git files in
1590 - some developer-related changes: improve the burnin and the QA suite,
1591 add an upload script for testing during development
1597 *(Released Wed, 16 Jan 2008)*
1599 - experimental HVM support, read the install document, section
1600 "Initializing the cluster"
1601 - allow for the PVM hypervisor per-instance kernel and initrd paths
1602 - add a new command ``gnt-cluster verify-disks`` which uses a new
1603 algorithm to improve the reconnection of the DRBD pairs if the device
1604 on the secondary node has gone away
1605 - make logical volume code auto-activate LVs at disk activation time
1606 - slightly improve the speed of activating disks
1607 - allow specification of the MAC address at instance creation time, and
1608 changing it later via ``gnt-instance modify``
1609 - fix handling of external commands that generate lots of output on
1611 - update documentation with regard to minimum version of DRBD8 supported
1617 *(Released Tue, 4 Dec 2007)*
1619 - Log the ``xm create`` output to the node daemon log on failure (to
1620 help diagnosing the error)
1621 - In debug mode, log all external commands output if failed to the logs
1622 - Change parsing of lvm commands to ignore stderr
1628 *(Released Wed, 28 Nov 2007)*
1630 - Another round of updates to the DRBD 8 code to deal with more failures
1631 in the replace secondary node operation
1632 - Some more logging of failures in disk operations (lvm, drbd)
1633 - A few documentation updates
1640 *(Released Tue, 13 Nov 2007)*
1642 - Change configuration file format from Python's Pickle to JSON.
1643 Upgrading is possible using the cfgupgrade utility.
1644 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1645 faster replace disks and is more stable (DRBD 8 has many improvements
1646 compared to DRBD 0.7)
1647 - Added command line tags support (see man pages for ``gnt-instance``,
1648 ``gnt-node``, ``gnt-cluster``)
1649 - Added instance rename support
1650 - Added multi-instance startup/shutdown
1651 - Added cluster rename support
1652 - Added ``gnt-node evacuate`` to simplify some node operations
1653 - Added instance reboot operation that can speedup reboot as compared to
1655 - Soften the requirement that hostnames are in FQDN format
1656 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1658 - Removed dependency on debian's patched fping that uses the
1659 non-standard ``-S`` option
1660 - Now the OS definitions are searched for in multiple, configurable
1661 paths (easier for distros to package)
1662 - Some changes to the hooks infrastructure (especially the new
1663 post-configuration update hook)
1664 - Other small bugfixes
1666 .. vim: set textwidth=72 syntax=rst :