7 *(Released Mon, 8 Aug 2011)*
9 Many bug-fixes and a few small features:
11 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
12 issues when you tried to add an instance with two MAC addresses in one
14 - KVM: fixed per-instance stored UID value
15 - KVM: configure bridged NICs at migration start
16 - KVM: Fix a bug where instance will not start with never KVM versions
18 - Added OS search path to ``gnt-cluster info``
19 - Fixed an issue with ``file_storage_dir`` where you were forced to
20 provide an absolute path, but the documentation states it is a
21 relative path, the documentation was right
22 - Added a new parameter to instance stop/start called ``--no-remember``
23 that will make the state change to not be remembered
24 - Implemented ``no_remember`` at RAPI level
25 - Improved the documentation
26 - Node evacuation: don't call IAllocator if node is already empty
27 - Fixed bug in DRBD8 replace disks on current nodes
28 - Fixed bug in recreate-disks for DRBD instances
29 - Moved assertion checking locks in ``gnt-instance replace-disks``
30 causing it to abort with not owning the right locks for some situation
31 - Job queue: Fixed potential race condition when cancelling queued jobs
32 - Fixed off-by-one bug in job serial generation
33 - ``gnt-node volumes``: Fix instance names
34 - Fixed aliases in bash completion
35 - Fixed a bug in reopening log files after beeing sent a SIGHUP
36 - Added a flag to burnin to allow specifying VCPU count
37 - Bugfixes to non-root Ganeti configuration
43 *(Released Thu, 12 May 2011)*
45 Many bug-fixes and a few new small features:
47 - Fixed a bug related to log opening failures
48 - Fixed a bug in instance listing with orphan instances
49 - Fixed a bug which prevented resetting the cluster-level node parameter
50 ``oob_program`` to the default
51 - Many fixes related to the ``cluster-merge`` tool
52 - Fixed a race condition in the lock monitor, which caused failures
53 during (at least) creation of many instances in parallel
54 - Improved output for gnt-job info
55 - Removed the quiet flag on some ssh calls which prevented debugging
57 - Improved the N+1 failure messages in cluster verify by actually
58 showing the memory values (needed and available)
59 - Increased lock attempt timeouts so that when executing long operations
60 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
61 too early and thus prevent the use of the 'fair' mechanism
62 - Changed instance query data (``gnt-instance info``) to not acquire
63 locks unless needed, thus allowing its use on locked instance if only
64 static information is asked for
65 - Improved behaviour with filesystems that do not support rename on an
67 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
68 kept locks on all nodes during the wipe, which is unneeded
69 - Fixed ``gnt-watcher`` handling of errors during hooks execution
70 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
71 10GiB) which caused the wipe to fail right at the end in some cases
72 - Fixed master IP activation when doing master failover with no-voting
73 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
74 the master node itself
75 - Fixed potential data-loss in under disk full conditions, where Ganeti
76 wouldn't check correctly the return code and would consider
77 partially-written files 'correct'
78 - Fixed bug related to multiple VGs and DRBD disk replacing
79 - Added new disk parameter ``metavg`` that allows placement of the meta
80 device for DRBD in a different volume group
81 - Fixed error handling in the node daemon when the system libc doesn't
82 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
83 - Fixed lock release during replace-disks, which kept cluster-wide locks
84 when doing disk replaces with an iallocator script
85 - Added check for missing bridges in cluster verify
86 - Handle EPIPE errors while writing to the terminal better, so that
87 piping the output to e.g. ``less`` doesn't cause a backtrace
88 - Fixed rare case where a ^C during Luxi calls could have been
89 interpreted as server errors, instead of simply terminating
90 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
92 - Added a few more parameters to the KVM hypervisor, allowing a second
93 CDROM, custom disk type for CDROMs and a floppy image
94 - Removed redundant message in instance rename when the name is given
96 - Added option to ``gnt-instance recreate-disks`` to allow creating the
97 disks on new nodes, allowing recreation when the original instance
98 nodes are completely gone
99 - Added option when converting disk templates to DRBD to skip waiting
100 for the resync, in order to make the instance available sooner
101 - Added two new variables to the OS scripts environment (containing the
103 - Made the root_path and optional parameter for the xen-pvm hypervisor,
104 to allow use of ``pvgrub`` as bootloader
105 - Changed the instance memory modifications to only check out-of-memory
106 conditions on memory increases, and turned the secondary node warnings
107 into errors (they can still be overridden via ``--force``)
108 - Fixed the handling of a corner case when the Python installation gets
109 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
110 to execute a command that doesn't exist
111 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
112 node of the instance returned failures during instance shutdown; this
113 adds the option ``--ignore-consistency`` to gnt-instance move
115 And as usual, various improvements to the error messages, documentation
121 *(Released Wed, 09 Mar 2011)*
123 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
124 the RAPI users file if run twice (even with ``--dry-run``).
126 The release fixes that bug (nothing else changed).
132 *(Released Mon, 07 Mar 2011)*
134 Final 2.4.0 release. Just a few small fixes:
136 - Fixed RAPI node evacuate
137 - Fixed the kvm-ifup script
138 - Fixed internal error handling for special job cases
139 - Updated man page to specify the escaping feature for options
145 *(Released Mon, 28 Feb 2011)*
147 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
148 that this feature wiped the disks of the wrong instance, leading to loss
153 - Fixed title of query field containing instance name
154 - Expanded the glossary in the documentation
155 - Fixed one unittest (internal issue)
161 *(Released Mon, 21 Feb 2011)*
163 A number of bug fixes plus just a couple functionality changes.
165 On the user-visible side, the ``gnt-* list`` command output has changed
166 with respect to "special" field states. The current rc1 style of display
167 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
168 the default output mode special fields states are displayed as follows:
170 - Offline resource: ``*``
171 - Unavailable/not applicable: ``-``
172 - Data missing (RPC failure): ``?``
173 - Unknown field: ``??``
175 Another user-visible change is the addition of ``--force-join`` to
180 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
181 - Fixed regression in RAPI/instance reinstall where all parameters were
182 required (instead of optional)
183 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
184 - Fixed iallocator usage (offline nodes were not considered offline)
185 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
186 - Fixed hypervisor and OS parameter validation with respect to
188 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
190 - Fixed ``tools/listrunner`` with respect to agent-based usage
196 *(Released Fri, 4 Feb 2011)*
198 Many changes and fixes since the beta1 release. While there were some
199 internal changes, the code has been mostly stabilised for the RC
202 Note: the dumb allocator was removed in this release, as it was not kept
203 up-to-date with the IAllocator protocol changes. It is recommended to
204 use the ``hail`` command from the ganeti-htools package.
206 Note: the 2.4 and up versions of Ganeti are not compatible with the
207 0.2.x branch of ganeti-htools. You need to upgrade to
208 ganeti-htools-0.3.0 (or later).
210 Regressions fixed from 2.3
211 ~~~~~~~~~~~~~~~~~~~~~~~~~~
213 - Fixed the ``gnt-cluster verify-disks`` command
214 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
216 - Fixed disk adoption breakage
217 - Fixed wrong headers in instance listing for field aliases
222 - Fixed corner case in KVM handling of NICs
223 - Fixed many cases of wrong handling of non-vm_capable nodes
224 - Fixed a bug where a missing instance symlink was not possible to
225 recreate with any ``gnt-*`` command (now ``gnt-instance
226 activate-disks`` does it)
227 - Fixed the volume group name as reported by ``gnt-cluster
229 - Increased timeouts for the import-export code, hopefully leading to
230 fewer aborts due network or instance timeouts
231 - Fixed bug in ``gnt-node list-storage``
232 - Fixed bug where not all daemons were started on cluster
233 initialisation, but only at the first watcher run
234 - Fixed many bugs in the OOB implementation
235 - Fixed watcher behaviour in presence of instances with offline
237 - Fixed instance list output for instances running on the wrong node
238 - a few fixes to the cluster-merge tool, but it still cannot merge
239 multi-node groups (currently it is not recommended to use this tool)
245 - Improved network configuration for the KVM hypervisor
246 - Added e1000 as a supported NIC for Xen-HVM
247 - Improved the lvmstrap tool to also be able to use partitions, as
248 opposed to full disks
249 - Improved speed of disk wiping (the cluster parameter
250 ``prealloc_wipe_disks``, so that it has a low impact on the total time
251 of instance creations
252 - Added documentation for the OS parameters
253 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
254 hypervisor is not responding
255 - Added display of blacklisted and hidden OS information in
257 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
258 NIC and node parameters, which might create problems with currently
259 invalid (but undetected) configuration files, but prevents validation
260 failures when unrelated parameters are modified
261 - Changed cluster initialisation to wait for the master daemon to become
263 - Expanded the RAPI interface:
265 - Added config redistribution resource
266 - Added activation/deactivation of instance disks
267 - Added export of console information
269 - Implemented log file reopening on SIGHUP, which allows using
270 logrotate(8) for the Ganeti log files
271 - Added a basic OOB helper script as an example
277 *(Released Fri, 14 Jan 2011)*
282 - Fixed timezone issues when formatting timestamps
283 - Added support for node groups, available via ``gnt-group`` and other
285 - Added out-of-band framework and management, see :doc:`design
286 document <design-oob>`
287 - Removed support for roman numbers from ``gnt-node list`` and
288 ``gnt-instance list``.
289 - Allowed modification of master network interface via ``gnt-cluster
290 modify --master-netdev``
291 - Accept offline secondaries while shutting down instance disks
292 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
293 - Added support for multiple LVM volume groups
294 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
296 - Added commands to list available fields:
298 - ``gnt-node list-fields``
299 - ``gnt-group list-fields``
300 - ``gnt-instance list-fields``
302 - Updated documentation and man pages
307 - Moved ``rapi_users`` file into separate directory, now named
308 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
310 - Added new tool for running commands on many machines,
311 ``tools/ganeti-listrunner``
312 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
313 improving the ``gnt-instance console`` output
314 - Allowed customisation of disk index separator at ``configure`` time
315 - Export node group allocation policy to :doc:`iallocator <iallocator>`
316 - Added support for non-partitioned md disks in ``lvmstrap``
317 - Added script to gracefully power off KVM instances
318 - Split ``utils`` module into smaller parts
319 - Changed query operations to return more detailed information, e.g.
320 whether an information is unavailable due to an offline node. To use
321 this new functionality, the LUXI call ``Query`` must be used. Field
322 information is now stored by the master daemon and can be retrieved
323 using ``QueryFields``. Instances, nodes and groups can also be queried
324 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
325 exposed via RAPI). The following commands make use of this
326 infrastructure change:
329 - ``gnt-group list-fields``
331 - ``gnt-node list-fields``
332 - ``gnt-instance list``
333 - ``gnt-instance list-fields``
334 - ``gnt-debug locks``
339 - New RAPI resources (see :doc:`rapi`):
343 - ``/2/groups/[group_name]``
344 - ``/2/groups/[group_name]/assign-nodes``
345 - ``/2/groups/[group_name]/modify``
346 - ``/2/groups/[group_name]/rename``
347 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
351 - Implemented ``no_install`` for instance creation
352 - Implemented OS parameters for instance reinstallation, allowing
353 use of special settings on reinstallation (e.g. for preserving data)
358 - Added IPv6 support in import/export
359 - Pause DRBD synchronization while wiping disks on instance creation
360 - Updated unittests and QA scripts
361 - Improved network parameters passed to KVM
362 - Converted man pages from docbook to reStructuredText
368 *(Released Mon, 20 Dec 2010)*
370 Released version 2.3.1~rc1 without any changes.
376 *(Released Wed, 1 Dec 2010)*
378 - impexpd: Disable OpenSSL compression in socat if possible (backport
379 from master, commit e90739d625b, see :doc:`installation guide
380 <install-quick>` for details)
381 - Changed unittest coverage report to exclude test scripts
382 - Added script to check version format
388 *(Released Wed, 1 Dec 2010)*
390 Released version 2.3.0~rc1 without any changes.
396 *(Released Fri, 19 Nov 2010)*
398 A number of bugfixes and documentation updates:
400 - Update ganeti-os-interface documentation
401 - Fixed a bug related to duplicate MACs or similar items which should be
403 - Fix breakage in OS state modify
404 - Reinstall instance: disallow offline secondaries (fixes bug related to
405 OS changing but reinstall failing)
406 - plus all the other fixes between 2.2.1 and 2.2.2
412 *(Released Tue, 2 Nov 2010)*
414 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
415 - Fixed master failover race with watcher
416 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
418 - Accept previously stopped instance for export with instance removal
419 - Simplify and extend the environment variables for instance OS scripts
420 - Added new node flags, ``master_capable`` and ``vm_capable``
421 - Added optional instance disk wiping prior during allocation. This is a
422 cluster-wide option and can be set/modified using
423 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
424 - Added IPv6 support, see :doc:`design document <design-2.3>` and
426 - Added a new watcher option (``--ignore-pause``)
427 - Added option to ignore offline node on instance start/stop
428 (``--ignore-offline``)
429 - Allow overriding OS parameters with ``gnt-instance reinstall``
430 - Added ability to change node's secondary IP address using ``gnt-node
432 - Implemented privilege separation for all daemons except
433 ``ganeti-noded``, see ``configure`` options
434 - Complain if an instance's disk is marked faulty in ``gnt-cluster
436 - Implemented job priorities (see ``ganeti(7)`` manpage)
437 - Ignore failures while shutting down instances during failover from
439 - Exit daemon's bootstrap process only once daemon is ready
440 - Export more information via ``LUInstanceQuery``/remote API
441 - Improved documentation, QA and unittests
442 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
443 restart if the file was created or changed
444 - Added LUXI protocol version sent with each request and response,
445 allowing detection of server/client mismatches
446 - Moved the Python scripts among gnt-* and ganeti-* into modules
447 - Moved all code related to setting up SSH to an external script,
449 - Infrastructure changes for node group support in future versions
455 *(Released Fri, 19 Nov 2010)*
457 A few small bugs fixed, and some improvements to the build system:
459 - Fix documentation regarding conversion to drbd
460 - Fix validation of parameters in cluster modify (``gnt-cluster modify
462 - Fix error handling in node modify with multiple changes
463 - Allow remote imports without checked names
469 *(Released Tue, 19 Oct 2010)*
471 - Disable SSL session ID cache in RPC client
477 *(Released Thu, 14 Oct 2010)*
479 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
480 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
482 - Fix problems with interaction between Curl and Python's HTTP server,
483 resulting in increased speed in many RPC calls
484 - Improve our release script to prevent breakage with older aclocal and
491 *(Released Thu, 7 Oct 2010)*
493 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
494 value retrieved from master
495 - Added support for blacklisted or hidden OS definitions
496 - Added simple lock monitor (accessible via (``gnt-debug locks``)
497 - Added support for -mem-path in KVM hypervisor abstraction layer
498 - Allow overriding instance parameters in tool for inter-cluster
499 instance moves (``tools/move-instance``)
500 - Improved opcode summaries (e.g. in ``gnt-job list``)
501 - Improve consistency of OS listing by sorting it
502 - Documentation updates
508 *(Released Fri, 8 Oct 2010)*
510 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
516 *(Released Mon, 4 Oct 2010)*
518 - Fixed regression in ``gnt-instance rename``
524 *(Released Wed, 22 Sep 2010)*
526 - Fixed OS_VARIANT variable for OS scripts
527 - Fixed cluster tag operations via RAPI
528 - Made ``setup-ssh`` exit with non-zero code if an error occurred
529 - Disabled RAPI CA checks in watcher
535 *(Released Mon, 23 Aug 2010)*
537 - Support DRBD versions of the format "a.b.c.d"
539 - Re-introduce support for usage from multiple threads in RAPI client
540 - Instance renames and modify via RAPI
541 - Work around race condition between processing and archival in job
543 - Mark opcodes following failed one as failed, too
544 - Job field ``lock_status`` was removed due to difficulties making it
545 work with the changed job queue in Ganeti 2.2; a better way to monitor
546 locks is expected for a later 2.2.x release
547 - Fixed dry-run behaviour with many commands
548 - Support ``ssh-agent`` again when adding nodes
549 - Many additional bugfixes
555 *(Released Fri, 30 Jul 2010)*
557 Important change: the internal RPC mechanism between Ganeti nodes has
558 changed from using a home-grown http library (based on the Python base
559 libraries) to use the PycURL library. This requires that PycURL is
560 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
561 against GnuTLS by default. cURL's support for GnuTLS had known issues
562 before cURL 7.21.0 and we recommend using the latest cURL release or
563 linking against OpenSSL. Most other distributions already link PycURL
564 and cURL against OpenSSL. The command::
566 python -c 'import pycurl; print pycurl.version'
568 can be used to determine the libraries PycURL and cURL are linked
571 Other significant changes:
573 - Rewrote much of the internals of the job queue, in order to achieve
574 better parallelism; this decouples job query operations from the job
575 processing, and it should allow much nicer behaviour of the master
576 daemon under load, and it also has uncovered some long-standing bugs
577 related to the job serialisation (now fixed)
578 - Added a default iallocator setting to the cluster parameters,
579 eliminating the need to always pass nodes or an iallocator for
580 operations that require selection of new node(s)
581 - Added experimental support for the LXC virtualization method
582 - Added support for OS parameters, which allows the installation of
583 instances to pass parameter to OS scripts in order to customise the
585 - Added a hypervisor parameter controlling the migration type (live or
586 non-live), since hypervisors have various levels of reliability; this
587 has renamed the 'live' parameter to 'mode'
588 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
589 logical volumes, meaning that cluster verify will ignore them and not
590 flag their presence as errors
591 - The watcher will now reset the error count for failed instances after
592 8 hours, thus allowing self-healing if the problem that caused the
593 instances to be down/fail to start has cleared in the meantime
594 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
595 check for, and warn, if the drbd module parameter ``usermode_helper``
596 is not consistent with the cluster-wide setting; this is needed to
597 make diagnose easier of failed drbd creations
598 - Started adding base IPv6 support, but this is not yet
599 enabled/available for use
600 - Rename operations (cluster, instance) will now return the new name,
601 which is especially useful if a short name was passed in
602 - Added support for instance migration in RAPI
603 - Added a tool to pre-configure nodes for the SSH setup, before joining
604 them to the cluster; this will allow in the future a simplified model
605 for node joining (but not yet fully enabled in 2.2); this needs the
606 paramiko python library
607 - Fixed handling of name-resolving errors
608 - Fixed consistency of job results on the error path
609 - Fixed master-failover race condition when executed multiple times in
611 - Fixed many bugs related to the job queue (mostly introduced during the
612 2.2 development cycle, so not all are impacting 2.1)
613 - Fixed instance migration with missing disk symlinks
614 - Fixed handling of unknown jobs in ``gnt-job archive``
615 - And many other small fixes/improvements
619 - Enhanced both the unittest and the QA coverage
620 - Switched the opcode validation to a generic model, and extended the
621 validation to all opcode parameters
622 - Changed more parts of the code that write shell scripts to use the
624 - Switched the master daemon to use the asyncore library for the Luxi
631 *(Released Thu, 17 Jun 2010)*
633 - Added tool (``move-instance``) and infrastructure to move instances
634 between separate clusters (see :doc:`separate documentation
635 <move-instance>` and :doc:`design document <design-2.2>`)
636 - Added per-request RPC timeout
637 - RAPI now requires a Content-Type header for requests with a body (e.g.
638 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
639 :rfc:`2616` (HTTP/1.1), section 7.2.1)
640 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
642 - Implemented initial support for running Ganeti daemons as separate
643 users, see configure-time flags ``--with-user-prefix`` and
644 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
646 - Instances can be removed after export (``gnt-backup export
648 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
649 key (instead of 1024 bit)
650 - Added new cluster configuration file for cluster domain secret
651 - Import/export now use SSL instead of SSH
652 - Added support for showing estimated time when exporting an instance,
653 see the ``ganeti-os-interface(7)`` manpage and look for
660 *(Released Tue, 16 Nov 2010)*
662 Some more bugfixes. Unless critical bugs occur, this will be the last
665 - Fix case of MAC special-values
666 - Fix mac checker regex
667 - backend: Fix typo causing "out of range" error
668 - Add missing --units in gnt-instance list man page
674 *(Released Tue, 24 Aug 2010)*
677 - Don't ignore secondary node silently on non-mirrored disk templates
679 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
680 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
681 - Properly document the usb_mouse parameter
682 - Fix path in ganeti-rapi(8) (issue 116)
683 - Adjust error message when the ganeti user's .ssh directory is
685 - Add same-node-check when changing the disk template to drbd
691 *(Released Fri, 16 Jul 2010)*
694 - Add an option to only select some reboot types during qa/burnin.
695 (on some hypervisors consequent reboots are not supported)
696 - Fix infrequent race condition in master failover. Sometimes the old
697 master ip address would be still detected as up for a short time
698 after it was removed, causing failover to fail.
699 - Decrease mlockall warnings when the ctypes module is missing. On
700 Python 2.4 we support running even if no ctypes module is installed,
701 but we were too verbose about this issue.
702 - Fix building on old distributions, on which man doesn't have a
704 - Fix RAPI not to ignore the MAC address on instance creation
705 - Implement the old instance creation format in the RAPI client.
711 *(Released Thu, 01 Jul 2010)*
713 A small bugfix release:
714 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
715 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
716 a non-existing option
717 - Fix instance create: the --force-variant option was ignored
718 - Improve pylint 0.21 compatibility and warnings with Python 2.6
719 - Fix modify node storage with non-FQDN arguments
720 - Fix RAPI client to authenticate under Python 2.6 when used
721 for more than 5 requests needing authentication
722 - Fix gnt-instance modify -t (storage) giving a wrong error message
723 when converting a non-shutdown drbd instance to plain
729 *(Released Fri, 18 Jun 2010)*
731 A small bugfix release:
733 - Fix live migration of KVM instances started with older Ganeti
734 versions which had fewer hypervisor parameters
735 - Fix gnt-instance grow-disk on down instances
736 - Fix an error-reporting bug during instance migration
737 - Better checking of the ``--net`` and ``--disk`` values, to avoid
738 silently ignoring broken ones
739 - Fix an RPC error reporting bug affecting, for example, RAPI client
741 - Fix bug triggered by different API version os-es on different nodes
742 - Fix a bug in instance startup with custom hvparams: OS level
743 parameters would fail to be applied.
744 - Fix the RAPI client under Python 2.6 (but more work is needed to
745 make it work completely well with OpenSSL)
746 - Fix handling of errors when resolving names from DNS
752 *(Released Thu, 3 Jun 2010)*
754 A medium sized development cycle. Some new features, and some
755 fixes/small improvements/cleanups.
760 The node deamon now tries to mlock itself into memory, unless the
761 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
762 its logs, and falls back to console logging. This allows emergency
763 features such as ``gnt-node powercycle`` to work even in the event of a
764 broken node disk (tested offlining the disk hosting the node's
765 filesystem and dropping its memory caches; don't try this at home)
767 KVM: add vhost-net acceleration support. It can be tested with a new
768 enough version of the kernel and of qemu-kvm.
770 KVM: Add instance chrooting feature. If you use privilege dropping for
771 your VMs you can also now force them to chroot to an empty directory,
772 before starting the emulated guest.
774 KVM: Add maximum migration bandwith and maximum downtime tweaking
775 support (requires a new-enough version of qemu-kvm).
777 Cluster verify will now warn if the master node doesn't have the master
780 Add a new (incompatible) instance creation request format to RAPI which
781 supports all parameters (previously only a subset was supported, and it
782 wasn't possible to extend the old format to accomodate all the new
783 features. The old format is still supported, and a client can check for
784 this feature, before using it, by checking for its presence in the
785 ``features`` RAPI resource.
787 Now with ancient latin support. Try it passing the ``--roman`` option to
788 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
789 (requires the python-roman module to be installed, in order to work).
794 As usual many internal code refactorings, documentation updates, and
797 - Lots of improvements and cleanups to the experimental Remote API
798 (RAPI) client library.
799 - A new unit test suite for the core daemon libraries.
800 - A fix to creating missing directories makes sure the umask is not
801 applied anymore. This enforces the same directory permissions
803 - Better handling terminating daemons with ctrl+c (used when running
804 them in debugging mode).
805 - Fix a race condition in live migrating a KVM instance, when stat()
806 on the old proc status file returned EINVAL, which is an unexpected
808 - Fixed manpage checking with newer man and utf-8 charachters. But now
809 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
815 *(Released Fri, 7 May 2010)*
817 Fix a bug which prevented untagged KVM instances from starting.
823 *(Released Fri, 7 May 2010)*
825 Another release with a long development cycle, during which many
826 different features were added.
831 The KVM hypervisor now can run the individual instances as non-root, to
832 reduce the impact of a VM being hijacked due to bugs in the
833 hypervisor. It is possible to run all instances as a single (non-root)
834 user, to manually specify a user for each instance, or to dynamically
835 allocate a user out of a cluster-wide pool to each instance, with the
836 guarantee that no two instances will run under the same user ID on any
839 An experimental RAPI client library, that can be used standalone
840 (without the other Ganeti libraries), is provided in the source tree as
841 ``lib/rapi/client.py``. Note this client might change its interface in
842 the future, as we iterate on its capabilities.
844 A new command, ``gnt-cluster renew-crypto`` has been added to easily
845 replace the cluster's certificates and crypto keys. This might help in
846 case they have been compromised, or have simply expired.
848 A new disk option for instance creation has been added that allows one
849 to "adopt" currently existing logical volumes, with data
850 preservation. This should allow easier migration to Ganeti from
851 unmanaged (or managed via other software) instances.
853 Another disk improvement is the possibility to convert between redundant
854 (DRBD) and plain (LVM) disk configuration for an instance. This should
855 allow better scalability (starting with one node and growing the
856 cluster, or shrinking a two-node cluster to one node).
858 A new feature that could help with automated node failovers has been
859 implemented: if a node sees itself as offline (by querying the master
860 candidates), it will try to shutdown (hard) all instances and any active
861 DRBD devices. This reduces the risk of duplicate instances if an
862 external script automatically failovers the instances on such nodes. To
863 enable this, the cluster parameter ``maintain_node_health`` should be
864 enabled; in the future this option (per the name) will enable other
865 automatic maintenance features.
867 Instance export/import now will reuse the original instance
868 specifications for all parameters; that means exporting an instance,
869 deleting it and the importing it back should give an almost identical
870 instance. Note that the default import behaviour has changed from
871 before, where it created only one NIC; now it recreates the original
874 Cluster verify has added a few new checks: SSL certificates validity,
875 /etc/hosts consistency across the cluster, etc.
880 As usual, many internal changes were done, documentation fixes,
883 - Fixed cluster initialization with disabled cluster storage (regression
885 - File-based storage supports growing the disks
886 - Fixed behaviour of node role changes
887 - Fixed cluster verify for some corner cases, plus a general rewrite of
888 cluster verify to allow future extension with more checks
889 - Fixed log spamming by watcher and node daemon (regression introduced
891 - Fixed possible validation issues when changing the list of enabled
893 - Fixed cleanup of /etc/hosts during node removal
894 - Fixed RAPI response for invalid methods
895 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
896 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
897 from ide disks, etc.)
898 - Allow OS changes without re-installation (to record a changed OS
899 outside of Ganeti, or to allow OS renames)
900 - Allow instance creation without OS installation (useful for example if
901 the OS will be installed manually, or restored from a backup not in
903 - Implemented option to make cluster ``copyfile`` use the replication
905 - Added list of enabled hypervisors to ssconf (possibly useful for
907 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
909 - A partial form of node re-IP is possible via node readd, which now
910 allows changed node primary IP
911 - Command line utilities now show an informational message if the job is
913 - The logs of the master daemon now show the PID/UID/GID of the
920 *(Released Fri, 12 Mar 2010)*
922 During the 2.1.0 long release candidate cycle, a lot of improvements and
923 changes have accumulated with were released later as 2.1.1.
928 The node evacuate command (``gnt-node evacuate``) was significantly
929 rewritten, and as such the IAllocator protocol was changed - a new
930 request type has been added. This unfortunate change during a stable
931 series is designed to improve performance of node evacuations; on
932 clusters with more than about five nodes and which are well-balanced,
933 evacuation should proceed in parallel for all instances of the node
934 being evacuated. As such, any existing IAllocator scripts need to be
935 updated, otherwise the above command will fail due to the unknown
936 request. The provided "dumb" allocator has not been updated; but the
937 ganeti-htools package supports the new protocol since version 0.2.4.
939 Another important change is increased validation of node and instance
940 names. This might create problems in special cases, if invalid host
941 names are being used.
943 Also, a new layer of hypervisor parameters has been added, that sits at
944 OS level between the cluster defaults and the instance ones. This allows
945 customisation of virtualization parameters depending on the installed
946 OS. For example instances with OS 'X' may have a different KVM kernel
947 (or any other parameter) than the cluster defaults. This is intended to
948 help managing a multiple OSes on the same cluster, without manual
949 modification of each instance's parameters.
951 A tool for merging clusters, ``cluster-merge``, has been added in the
957 - Improved the int/float conversions that should make the code more
958 robust in face of errors from the node daemons
959 - Fixed the remove node code in case of internal configuration errors
960 - Fixed the node daemon behaviour in face of inconsistent queue
961 directory (e.g. read-only file-system where we can't open the files
963 - Fixed the behaviour of gnt-node modify for master candidate demotion;
964 now it either aborts cleanly or, if given the new "auto_promote"
965 parameter, will automatically promote other nodes as needed
966 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
967 completely prevent Ganeti from working
968 - Fixed bug for instance export when not all disks were successfully
970 - Fixed behaviour of node add when the new node is slow in starting up
972 - Fixed handling of signals in the LUXI client, which should improve
973 behaviour of command-line scripts
974 - Added checks for invalid node/instance names in the configuration (now
975 flagged during cluster verify)
976 - Fixed watcher behaviour for disk activation errors
977 - Fixed two potentially endless loops in http library, which led to the
978 RAPI daemon hanging and consuming 100% CPU in some cases
979 - Fixed bug in RAPI daemon related to hashed passwords
980 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
982 - Enhanced compatibility with non-Debian OSes, but not using absolute
983 path in some commands and allowing customisation of the ssh
984 configuration directory
985 - Fixed possible future issue with new Python versions by abiding to the
986 proper use of ``__slots__`` attribute on classes
987 - Added checks that should prevent directory traversal attacks
988 - Many documentation fixes based on feedback from users
993 - Added an "early_release" more for instance replace disks and node
994 evacuate, where we release locks earlier and thus allow higher
995 parallelism within the cluster
996 - Added watcher hooks, intended to allow the watcher to restart other
997 daemons (e.g. from the ganeti-nbma project), but they can be used of
998 course for any other purpose
999 - Added a compile-time disable for DRBD barriers, to increase
1000 performance if the administrator trusts the power supply or the
1001 storage system to not lose writes
1002 - Added the option of using syslog for logging instead of, or in
1003 addition to, Ganeti's own log files
1004 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1005 can indeed boot from a paravirtual NIC
1006 - Added a generic debug level for many operations; while this is not
1007 used widely yet, it allows one to pass the debug value all the way to
1009 - Enhanced the hooks environment for instance moves (failovers,
1010 migrations) where the primary/secondary nodes changed during the
1011 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1012 - Enhanced data validations for many user-supplied values; one important
1013 item is the restrictions imposed on instance and node names, which
1014 might reject some (invalid) host names
1015 - Add a configure-time option to disable file-based storage, if it's not
1016 needed; this allows greater security separation between the master
1017 node and the other nodes from the point of view of the inter-node RPC
1019 - Added user notification in interactive tools if job is waiting in the
1020 job queue or trying to acquire locks
1021 - Added log messages when a job is waiting for locks
1022 - Added filtering by node tags in instance operations which admit
1023 multiple instances (start, stop, reboot, reinstall)
1024 - Added a new tool for cluster mergers, ``cluster-merge``
1025 - Parameters from command line which are of the form ``a=b,c=d`` can now
1026 use backslash escapes to pass in values which contain commas,
1027 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1029 - For KVM, the instance name is the first parameter passed to KVM, so
1030 that it's more visible in the process list
1036 *(Released Tue, 2 Mar 2010)*
1038 Ganeti 2.1 brings many improvements with it. Major changes:
1040 - Added infrastructure to ease automated disk repairs
1041 - Added new daemon to export configuration data in a cheaper way than
1042 using the remote API
1043 - Instance NICs can now be routed instead of being associated with a
1045 - Improved job locking logic to reduce impact of jobs acquiring multiple
1046 locks waiting for other long-running jobs
1048 In-depth implementation details can be found in the Ganeti 2.1 design
1054 - Added chroot hypervisor
1055 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1057 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1058 ``bootloader_path`` and ``bootloader_args``)
1059 - Added the ``use_localtime`` option for the xen-hvm and kvm
1060 hypervisors, and the default value for this has changed to false (in
1061 2.0 xen-hvm always enabled it)
1062 - Added luxi call to submit multiple jobs in one go
1063 - Added cluster initialization option to not modify ``/etc/hosts``
1065 - Added network interface parameters
1066 - Added dry run mode to some LUs
1067 - Added RAPI resources:
1069 - ``/2/instances/[instance_name]/info``
1070 - ``/2/instances/[instance_name]/replace-disks``
1071 - ``/2/nodes/[node_name]/evacuate``
1072 - ``/2/nodes/[node_name]/migrate``
1073 - ``/2/nodes/[node_name]/role``
1074 - ``/2/nodes/[node_name]/storage``
1075 - ``/2/nodes/[node_name]/storage/modify``
1076 - ``/2/nodes/[node_name]/storage/repair``
1078 - Added OpCodes to evacuate or migrate all instances on a node
1079 - Added new command to list storage elements on nodes (``gnt-node
1080 list-storage``) and modify them (``gnt-node modify-storage``)
1081 - Added new ssconf files with master candidate IP address
1082 (``ssconf_master_candidates_ips``), node primary IP address
1083 (``ssconf_node_primary_ips``) and node secondary IP address
1084 (``ssconf_node_secondary_ips``)
1085 - Added ``ganeti-confd`` and a client library to query the Ganeti
1086 configuration via UDP
1087 - Added ability to run hooks after cluster initialization and before
1089 - Added automatic mode for disk replace (``gnt-instance replace-disks
1091 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1092 after catastrophic data-loss
1093 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1095 - Added ``gnt-instance move`` command to move instances
1096 - Added ``gnt-cluster watcher`` command to control watcher
1097 - Added ``gnt-node powercycle`` command to powercycle nodes
1098 - Added new job status field ``lock_status``
1099 - Added parseable error codes to cluster verification (``gnt-cluster
1100 verify --error-codes``) and made output less verbose (use
1101 ``--verbose`` to restore previous behaviour)
1102 - Added UUIDs to the main config entities (cluster, nodes, instances)
1103 - Added support for OS variants
1104 - Added support for hashed passwords in the Ganeti remote API users file
1106 - Added option to specify maximum timeout on instance shutdown
1107 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1108 - Added new helper script to start and stop Ganeti daemons
1109 (``daemon-util``), with the intent to reduce the work necessary to
1110 adjust Ganeti for non-Debian distributions and to start/stop daemons
1112 - Added more unittests
1113 - Fixed critical bug in ganeti-masterd startup
1114 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1115 now customisable at the cluster level for both the KVM and Xen
1116 hypervisors using the new ``migration_port`` parameter
1117 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1118 reinstalling an instance
1119 - Allowed ``@`` in tag names
1120 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1121 - Many documentation updates
1122 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1123 - ``gnt-instance reinstall`` can now reinstall multiple instances
1124 - Updated many command line parameters
1125 - Introduced new OS API version 15
1126 - No longer support a default hypervisor
1127 - Treat virtual LVs as inexistent
1128 - Improved job locking logic to reduce lock contention
1129 - Match instance and node names case insensitively
1130 - Reimplemented bash completion script to be more complete
1137 *(Released Thu, 4 Feb 2010)*
1139 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1140 - Fix a string formatting bug
1141 - Improve safety of the code in some error paths
1142 - Improve data validation in the master of values returned from nodes
1148 *(Released Thu, 17 Dec 2009)*
1150 - Fix security issue due to missing validation of iallocator names; this
1151 allows local and remote execution of arbitrary executables
1152 - Fix failure of gnt-node list during instance removal
1153 - Ship the RAPI documentation in the archive
1159 *(Released Wed, 30 Sep 2009)*
1161 - Fixed many wrong messages
1162 - Fixed a few bugs related to the locking library
1163 - Fixed MAC checking at instance creation time
1164 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1165 - Fixed a few issues related to signal handling in both daemons and
1167 - Fixed the example startup script provided
1168 - Fixed insserv dependencies in the example startup script (patch from
1170 - Fixed handling of drained nodes in the iallocator framework
1171 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1173 - Fixed error related to invalid job IDs in job polling
1174 - Fixed job/opcode persistence on unclean master shutdown
1175 - Fixed handling of partial job processing after unclean master
1177 - Fixed error reporting from LUs, previously all errors were converted
1178 into execution errors
1179 - Fixed error reporting from burnin
1180 - Decreased significantly the memory usage of the job queue
1181 - Optimised slightly multi-job submission
1182 - Optimised slightly opcode loading
1183 - Backported the multi-job submit framework from the development
1184 branch; multi-instance start and stop should be faster
1185 - Added script to clean archived jobs after 21 days; this will reduce
1186 the size of the queue directory
1187 - Added some extra checks in disk size tracking
1188 - Added an example ethers hook script
1189 - Added a cluster parameter that prevents Ganeti from modifying of
1191 - Added more node information to RAPI responses
1192 - Added a ``gnt-job watch`` command that allows following the ouput of a
1194 - Added a bind-address option to ganeti-rapi
1195 - Added more checks to the configuration verify
1196 - Enhanced the burnin script such that some operations can be retried
1198 - Converted instance reinstall to multi-instance model
1204 *(Released Fri, 7 Aug 2009)*
1206 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1207 to allow using the pre-2.0.2 behaviour in activation, if any existing
1208 instances have mismatched disk sizes in the configuration
1209 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1210 any configuration mismatches for disk sizes
1211 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1212 failover to work on two-node clusters
1213 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1215 - Fixed detection of OS script errors in ``gnt-backup export``
1216 - Fixed exit code of ``gnt-backup export``
1222 *(Released Fri, 17 Jul 2009)*
1224 - Added experimental support for stripped logical volumes; this should
1225 enhance performance but comes with a higher complexity in the block
1226 device handling; stripping is only enabled when passing
1227 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1228 affected even in the non-stripped mode
1229 - Improved resiliency against transient failures at the end of DRBD
1230 resyncs, and in general of DRBD resync checks
1231 - Fixed a couple of issues with exports and snapshot errors
1232 - Fixed a couple of issues in instance listing
1233 - Added display of the disk size in ``gnt-instance info``
1234 - Fixed checking for valid OSes in instance creation
1235 - Fixed handling of the "vcpus" parameter in instance listing and in
1236 general of invalid parameters
1237 - Fixed http server library, and thus RAPI, to handle invalid
1238 username/password combinations correctly; this means that now they
1239 report unauthorized for queries too, not only for modifications,
1240 allowing earlier detect of configuration problems
1241 - Added a new "role" node list field, equivalent to the master/master
1242 candidate/drained/offline flags combinations
1243 - Fixed cluster modify and changes of candidate pool size
1244 - Fixed cluster verify error messages for wrong files on regular nodes
1245 - Fixed a couple of issues with node demotion from master candidate role
1246 - Fixed node readd issues
1247 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1248 - Added a new ``--no-voting`` option for masterfailover to fix failover
1249 on two-nodes clusters when the former master node is unreachable
1250 - Added instance reinstall over RAPI
1256 *(Released Tue, 16 Jun 2009)*
1258 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1259 allow re-adding the start in single-user option (regression from 1.2)
1260 - the watcher writes the instance status to a file, to allow monitoring
1261 to report the instance status (from the master) based on cached
1262 results of the watcher's queries; while this can get stale if the
1263 watcher is being locked due to other work on the cluster, this is
1264 still an improvement
1265 - the watcher now also restarts the node daemon and the rapi daemon if
1267 - fixed the watcher to handle full and drained queue cases
1268 - hooks export more instance data in the environment, which helps if
1269 hook scripts need to take action based on the instance's properties
1270 (no longer need to query back into ganeti)
1271 - instance failovers when the instance is stopped do not check for free
1272 RAM, so that failing over a stopped instance is possible in low memory
1274 - rapi uses queries for tags instead of jobs (for less job traffic), and
1275 for cluster tags it won't talk to masterd at all but read them from
1277 - a couple of error handling fixes in RAPI
1278 - drbd handling: improved the error handling of inconsistent disks after
1279 resync to reduce the frequency of "there are some degraded disks for
1280 this instance" messages
1281 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1282 error handling path called a wrong function name)
1288 *(Released Wed, 27 May 2009)*
1290 - no changes from rc5
1293 Version 2.0 release candidate 5
1294 -------------------------------
1296 *(Released Wed, 20 May 2009)*
1298 - fix a couple of bugs (validation, argument checks)
1299 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1300 - some small improvements to RAPI and IAllocator
1301 - make watcher automatically start the master daemon if down
1304 Version 2.0 release candidate 4
1305 -------------------------------
1307 *(Released Mon, 27 Apr 2009)*
1309 - change the OS list to not require locks; this helps with big clusters
1310 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1311 volume group is broken
1312 - ``gnt-instance info``, without any arguments, doesn't run for all
1313 instances anymore; either pass ``--all`` or pass the desired
1314 instances; this helps against mistakes on big clusters where listing
1315 the information for all instances takes a long time
1316 - miscellaneous doc and man pages fixes
1319 Version 2.0 release candidate 3
1320 -------------------------------
1322 *(Released Wed, 8 Apr 2009)*
1324 - Change the internal locking model of some ``gnt-node`` commands, in
1325 order to reduce contention (and blocking of master daemon) when
1326 batching many creation/reinstall jobs
1327 - Fixes to Xen soft reboot
1328 - No longer build documentation at build time, instead distribute it in
1329 the archive, in order to reduce the need for the whole docbook/rst
1333 Version 2.0 release candidate 2
1334 -------------------------------
1336 *(Released Fri, 27 Mar 2009)*
1338 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1339 - Fix watcher startup sequence, improves the behaviour of busy clusters
1340 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1341 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1342 bind address and other places
1343 - Some documentation fixes and updates
1346 Version 2.0 release candidate 1
1347 -------------------------------
1349 *(Released Mon, 2 Mar 2009)*
1351 - More documentation updates, now all docs should be more-or-less
1353 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1355 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1357 - Fix an issue related to $libdir/run/ganeti and cluster creation
1363 *(Released Thu, 19 Feb 2009)*
1365 - Xen PVM and KVM have switched the default value for the instance root
1366 disk to the first partition on the first drive, instead of the whole
1367 drive; this means that the OS installation scripts must be changed
1369 - Man pages have been updated
1370 - RAPI has been switched by default to HTTPS, and the exported functions
1371 should all work correctly
1372 - RAPI v1 has been removed
1373 - Many improvements to the KVM hypervisor
1374 - Block device errors are now better reported
1375 - Many other bugfixes and small improvements
1381 *(Released Mon, 26 Jan 2009)*
1383 - Version 2 is a general rewrite of the code and therefore the
1384 differences are too many to list, see the design document for 2.0 in
1385 the ``doc/`` subdirectory for more details
1386 - In this beta version there is not yet a migration path from 1.2 (there
1387 will be one in the final 2.0 release)
1388 - A few significant changes are:
1390 - all commands are executed by a daemon (``ganeti-masterd``) and the
1391 various ``gnt-*`` commands are just front-ends to it
1392 - all the commands are entered into, and executed from a job queue,
1393 see the ``gnt-job(8)`` manpage
1394 - the RAPI daemon supports read-write operations, secured by basic
1395 HTTP authentication on top of HTTPS
1396 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1397 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1398 to migrate to DRBD 8 first while still running Ganeti 1.2)
1399 - DRBD devices are using statically allocated minor numbers, which
1400 will be assigned to existing instances during the migration process
1401 - there is support for both Xen PVM and Xen HVM instances running on
1403 - KVM virtualization is supported too
1404 - file-based storage has been implemented, which means that it is
1405 possible to run the cluster without LVM and DRBD storage, for
1406 example using a shared filesystem exported from shared storage (and
1407 still have live migration)
1413 *(Released Tue, 13 Jan 2009)*
1415 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1416 - Reuse the old instance mac address by default on instance import, if
1417 the instance name is the same.
1418 - Handle situations in which the node info rpc returns incomplete
1420 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1421 - Improved version of batcher:
1423 - state file support
1424 - instance mac address support
1425 - support for HVM clusters/instances
1427 - Add an option to show the number of cpu sockets and nodes in
1429 - Support OSes that handle more than one version of the OS api (but do
1430 not change the current API in any other way)
1431 - Fix ``gnt-node migrate``
1432 - ``gnt-debug`` man page
1433 - Fixes various more typos and small issues
1434 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1440 *(Released Wed, 24 Sep 2008)*
1442 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1443 type of disk exported to fully virtualized instances.
1444 - provide access to the serial console of HVM instances
1445 - instance auto_balance flag, set by default. If turned off it will
1446 avoid warnings on cluster verify if there is not enough memory to fail
1447 over an instance. in the future it will prevent automatically failing
1448 it over when we will support that.
1449 - batcher tool for instance creation, see ``tools/README.batcher``
1450 - ``gnt-instance reinstall --select-os`` to interactively select a new
1451 operating system when reinstalling an instance.
1452 - when changing the memory amount on instance modify a check has been
1453 added that the instance will be able to start. also warnings are
1454 emitted if the instance will not be able to fail over, if auto_balance
1456 - documentation fixes
1457 - sync fields between ``gnt-instance list/modify/add/import``
1458 - fix a race condition in drbd when the sync speed was set after giving
1459 the device a remote peer.
1465 *(Released Tue, 22 Jul 2008)*
1467 - note: the allowed size and number of tags per object were reduced
1468 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1469 - fixed twisted 8.x compatibility
1470 - fixed ``gnt-instance replace-disks`` with iallocator
1471 - add TCP keepalives on twisted connections to detect restarted nodes
1472 - disk increase support, see ``gnt-instance grow-disk``
1473 - implement bulk node/instance query for RAPI
1474 - add tags in node/instance listing (optional)
1475 - experimental migration (and live migration) support, read the man page
1476 for ``gnt-instance migrate``
1477 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1478 has some small improvements in handling its state file
1484 *(Released Fri, 13 Jun 2008)*
1486 - Experimental readonly, REST-based remote API implementation;
1487 automatically started on master node, TCP port 5080, if enabled by
1488 ``--enable-rapi`` parameter to configure script.
1489 - Instance allocator support. Add and import instance accept a
1490 ``--iallocator`` parameter, and call that instance allocator to decide
1491 which node to use for the instance. The iallocator document describes
1492 what's expected from an allocator script.
1493 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1494 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1495 node is lost there is still enough memory to fail over the instances
1497 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1498 ``gnt-cluster verify``, to check for site-specific compliance. All the
1499 hooks will run, and their output, if any, will be displayed. Any
1500 failing hook will make the verification return an error value.
1501 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1502 primary and secondary interfaces
1503 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1504 that is still declared as part of the cluster and has failed.
1505 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1506 specifying output fields, that just adds the chosen fields to the
1508 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1509 export from the filesystem.
1510 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1511 have been added. Using them you can enable/disable acpi and pae
1512 support, and specify a path for a cd image to be exported to the
1513 instance. These parameters as the name suggest only work on HVM
1515 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1516 PAE support will be set to the previously hardcoded values, but the
1517 (previously hardcoded) path to the CDROM ISO image will be unset and
1518 if required, needs to be set manually with ``gnt-instance modify``
1520 - The address to which an instance's VNC console is bound is now
1521 selectable per-instance, rather than being cluster wide. Of course
1522 this only applies to instances controlled via VNC, so currently just
1523 applies to HVM clusters.
1529 *(Released Mon, 18 Feb 2008)*
1531 - more tweaks to the disk activation code (especially helpful for DRBD)
1532 - change the default ``gnt-instance list`` output format, now there is
1533 one combined status field (see the manpage for the exact values this
1535 - some more fixes for the mac export to hooks change
1536 - make Ganeti not break with DRBD 8.2.x (which changed the version
1537 format in ``/proc/drbd``) (issue 24)
1538 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1539 template, allowing migration from DRBD0.7+MD to DRBD8
1545 *(Released Wed, 30 Jan 2008)*
1547 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1549 - add command aliases infrastructure and a few aliases
1550 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1551 man pages and the ``--help`` option of ``gnt-node
1552 list``/``gnt-instance list``
1553 - fix ``gnt-backup list`` with down nodes (issue 21)
1554 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1555 - fix the dist archive and add a check for including svn/git files in
1557 - some developer-related changes: improve the burnin and the QA suite,
1558 add an upload script for testing during development
1564 *(Released Wed, 16 Jan 2008)*
1566 - experimental HVM support, read the install document, section
1567 "Initializing the cluster"
1568 - allow for the PVM hypervisor per-instance kernel and initrd paths
1569 - add a new command ``gnt-cluster verify-disks`` which uses a new
1570 algorithm to improve the reconnection of the DRBD pairs if the device
1571 on the secondary node has gone away
1572 - make logical volume code auto-activate LVs at disk activation time
1573 - slightly improve the speed of activating disks
1574 - allow specification of the MAC address at instance creation time, and
1575 changing it later via ``gnt-instance modify``
1576 - fix handling of external commands that generate lots of output on
1578 - update documentation with regard to minimum version of DRBD8 supported
1584 *(Released Tue, 4 Dec 2007)*
1586 - Log the ``xm create`` output to the node daemon log on failure (to
1587 help diagnosing the error)
1588 - In debug mode, log all external commands output if failed to the logs
1589 - Change parsing of lvm commands to ignore stderr
1595 *(Released Wed, 28 Nov 2007)*
1597 - Another round of updates to the DRBD 8 code to deal with more failures
1598 in the replace secondary node operation
1599 - Some more logging of failures in disk operations (lvm, drbd)
1600 - A few documentation updates
1607 *(Released Tue, 13 Nov 2007)*
1609 - Change configuration file format from Python's Pickle to JSON.
1610 Upgrading is possible using the cfgupgrade utility.
1611 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1612 faster replace disks and is more stable (DRBD 8 has many improvements
1613 compared to DRBD 0.7)
1614 - Added command line tags support (see man pages for ``gnt-instance``,
1615 ``gnt-node``, ``gnt-cluster``)
1616 - Added instance rename support
1617 - Added multi-instance startup/shutdown
1618 - Added cluster rename support
1619 - Added ``gnt-node evacuate`` to simplify some node operations
1620 - Added instance reboot operation that can speedup reboot as compared to
1622 - Soften the requirement that hostnames are in FQDN format
1623 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1625 - Removed dependency on debian's patched fping that uses the
1626 non-standard ``-S`` option
1627 - Now the OS definitions are searched for in multiple, configurable
1628 paths (easier for distros to package)
1629 - Some changes to the hooks infrastructure (especially the new
1630 post-configuration update hook)
1631 - Other small bugfixes
1633 .. vim: set textwidth=72 syntax=rst :