7 *(Released Thu, 27 Oct 2011)*
9 - Fixed bug when parsing command line parameter values ending in
11 - Fixed assertion error after unclean master shutdown
12 - Disable HTTP client pool for RPC, significantly reducing memory usage
14 - Fixed queue archive creation with wrong permissions
20 *(Released Tue, 23 Aug 2011)*
24 - Fixed documentation for importing with ``--src-dir`` option
25 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
26 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
32 *(Released Fri, 5 Aug 2011)*
34 Many bug-fixes and a few small features:
36 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
37 issues when you tried to add an instance with two MAC addresses in one
39 - KVM: fixed per-instance stored UID value
40 - KVM: configure bridged NICs at migration start
41 - KVM: Fix a bug where instance will not start with never KVM versions
43 - Added OS search path to ``gnt-cluster info``
44 - Fixed an issue with ``file_storage_dir`` where you were forced to
45 provide an absolute path, but the documentation states it is a
46 relative path, the documentation was right
47 - Added a new parameter to instance stop/start called ``--no-remember``
48 that will make the state change to not be remembered
49 - Implemented ``no_remember`` at RAPI level
50 - Improved the documentation
51 - Node evacuation: don't call IAllocator if node is already empty
52 - Fixed bug in DRBD8 replace disks on current nodes
53 - Fixed bug in recreate-disks for DRBD instances
54 - Moved assertion checking locks in ``gnt-instance replace-disks``
55 causing it to abort with not owning the right locks for some situation
56 - Job queue: Fixed potential race condition when cancelling queued jobs
57 - Fixed off-by-one bug in job serial generation
58 - ``gnt-node volumes``: Fix instance names
59 - Fixed aliases in bash completion
60 - Fixed a bug in reopening log files after being sent a SIGHUP
61 - Added a flag to burnin to allow specifying VCPU count
62 - Bugfixes to non-root Ganeti configuration
68 *(Released Thu, 12 May 2011)*
70 Many bug-fixes and a few new small features:
72 - Fixed a bug related to log opening failures
73 - Fixed a bug in instance listing with orphan instances
74 - Fixed a bug which prevented resetting the cluster-level node parameter
75 ``oob_program`` to the default
76 - Many fixes related to the ``cluster-merge`` tool
77 - Fixed a race condition in the lock monitor, which caused failures
78 during (at least) creation of many instances in parallel
79 - Improved output for gnt-job info
80 - Removed the quiet flag on some ssh calls which prevented debugging
82 - Improved the N+1 failure messages in cluster verify by actually
83 showing the memory values (needed and available)
84 - Increased lock attempt timeouts so that when executing long operations
85 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
86 too early and thus prevent the use of the 'fair' mechanism
87 - Changed instance query data (``gnt-instance info``) to not acquire
88 locks unless needed, thus allowing its use on locked instance if only
89 static information is asked for
90 - Improved behaviour with filesystems that do not support rename on an
92 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
93 kept locks on all nodes during the wipe, which is unneeded
94 - Fixed ``gnt-watcher`` handling of errors during hooks execution
95 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
96 10GiB) which caused the wipe to fail right at the end in some cases
97 - Fixed master IP activation when doing master failover with no-voting
98 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
99 the master node itself
100 - Fixed potential data-loss in under disk full conditions, where Ganeti
101 wouldn't check correctly the return code and would consider
102 partially-written files 'correct'
103 - Fixed bug related to multiple VGs and DRBD disk replacing
104 - Added new disk parameter ``metavg`` that allows placement of the meta
105 device for DRBD in a different volume group
106 - Fixed error handling in the node daemon when the system libc doesn't
107 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
108 - Fixed lock release during replace-disks, which kept cluster-wide locks
109 when doing disk replaces with an iallocator script
110 - Added check for missing bridges in cluster verify
111 - Handle EPIPE errors while writing to the terminal better, so that
112 piping the output to e.g. ``less`` doesn't cause a backtrace
113 - Fixed rare case where a ^C during Luxi calls could have been
114 interpreted as server errors, instead of simply terminating
115 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
117 - Added a few more parameters to the KVM hypervisor, allowing a second
118 CDROM, custom disk type for CDROMs and a floppy image
119 - Removed redundant message in instance rename when the name is given
121 - Added option to ``gnt-instance recreate-disks`` to allow creating the
122 disks on new nodes, allowing recreation when the original instance
123 nodes are completely gone
124 - Added option when converting disk templates to DRBD to skip waiting
125 for the resync, in order to make the instance available sooner
126 - Added two new variables to the OS scripts environment (containing the
128 - Made the root_path and optional parameter for the xen-pvm hypervisor,
129 to allow use of ``pvgrub`` as bootloader
130 - Changed the instance memory modifications to only check out-of-memory
131 conditions on memory increases, and turned the secondary node warnings
132 into errors (they can still be overridden via ``--force``)
133 - Fixed the handling of a corner case when the Python installation gets
134 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
135 to execute a command that doesn't exist
136 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
137 node of the instance returned failures during instance shutdown; this
138 adds the option ``--ignore-consistency`` to gnt-instance move
140 And as usual, various improvements to the error messages, documentation
146 *(Released Wed, 09 Mar 2011)*
148 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
149 the RAPI users file if run twice (even with ``--dry-run``).
151 The release fixes that bug (nothing else changed).
157 *(Released Mon, 07 Mar 2011)*
159 Final 2.4.0 release. Just a few small fixes:
161 - Fixed RAPI node evacuate
162 - Fixed the kvm-ifup script
163 - Fixed internal error handling for special job cases
164 - Updated man page to specify the escaping feature for options
170 *(Released Mon, 28 Feb 2011)*
172 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
173 that this feature wiped the disks of the wrong instance, leading to loss
178 - Fixed title of query field containing instance name
179 - Expanded the glossary in the documentation
180 - Fixed one unittest (internal issue)
186 *(Released Mon, 21 Feb 2011)*
188 A number of bug fixes plus just a couple functionality changes.
190 On the user-visible side, the ``gnt-* list`` command output has changed
191 with respect to "special" field states. The current rc1 style of display
192 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
193 the default output mode special fields states are displayed as follows:
195 - Offline resource: ``*``
196 - Unavailable/not applicable: ``-``
197 - Data missing (RPC failure): ``?``
198 - Unknown field: ``??``
200 Another user-visible change is the addition of ``--force-join`` to
205 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
206 - Fixed regression in RAPI/instance reinstall where all parameters were
207 required (instead of optional)
208 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
209 - Fixed iallocator usage (offline nodes were not considered offline)
210 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
211 - Fixed hypervisor and OS parameter validation with respect to
213 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
215 - Fixed ``tools/listrunner`` with respect to agent-based usage
221 *(Released Fri, 4 Feb 2011)*
223 Many changes and fixes since the beta1 release. While there were some
224 internal changes, the code has been mostly stabilised for the RC
227 Note: the dumb allocator was removed in this release, as it was not kept
228 up-to-date with the IAllocator protocol changes. It is recommended to
229 use the ``hail`` command from the ganeti-htools package.
231 Note: the 2.4 and up versions of Ganeti are not compatible with the
232 0.2.x branch of ganeti-htools. You need to upgrade to
233 ganeti-htools-0.3.0 (or later).
235 Regressions fixed from 2.3
236 ~~~~~~~~~~~~~~~~~~~~~~~~~~
238 - Fixed the ``gnt-cluster verify-disks`` command
239 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
241 - Fixed disk adoption breakage
242 - Fixed wrong headers in instance listing for field aliases
247 - Fixed corner case in KVM handling of NICs
248 - Fixed many cases of wrong handling of non-vm_capable nodes
249 - Fixed a bug where a missing instance symlink was not possible to
250 recreate with any ``gnt-*`` command (now ``gnt-instance
251 activate-disks`` does it)
252 - Fixed the volume group name as reported by ``gnt-cluster
254 - Increased timeouts for the import-export code, hopefully leading to
255 fewer aborts due network or instance timeouts
256 - Fixed bug in ``gnt-node list-storage``
257 - Fixed bug where not all daemons were started on cluster
258 initialisation, but only at the first watcher run
259 - Fixed many bugs in the OOB implementation
260 - Fixed watcher behaviour in presence of instances with offline
262 - Fixed instance list output for instances running on the wrong node
263 - a few fixes to the cluster-merge tool, but it still cannot merge
264 multi-node groups (currently it is not recommended to use this tool)
270 - Improved network configuration for the KVM hypervisor
271 - Added e1000 as a supported NIC for Xen-HVM
272 - Improved the lvmstrap tool to also be able to use partitions, as
273 opposed to full disks
274 - Improved speed of disk wiping (the cluster parameter
275 ``prealloc_wipe_disks``, so that it has a low impact on the total time
276 of instance creations
277 - Added documentation for the OS parameters
278 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
279 hypervisor is not responding
280 - Added display of blacklisted and hidden OS information in
282 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
283 NIC and node parameters, which might create problems with currently
284 invalid (but undetected) configuration files, but prevents validation
285 failures when unrelated parameters are modified
286 - Changed cluster initialisation to wait for the master daemon to become
288 - Expanded the RAPI interface:
290 - Added config redistribution resource
291 - Added activation/deactivation of instance disks
292 - Added export of console information
294 - Implemented log file reopening on SIGHUP, which allows using
295 logrotate(8) for the Ganeti log files
296 - Added a basic OOB helper script as an example
302 *(Released Fri, 14 Jan 2011)*
307 - Fixed timezone issues when formatting timestamps
308 - Added support for node groups, available via ``gnt-group`` and other
310 - Added out-of-band framework and management, see :doc:`design
311 document <design-oob>`
312 - Removed support for roman numbers from ``gnt-node list`` and
313 ``gnt-instance list``.
314 - Allowed modification of master network interface via ``gnt-cluster
315 modify --master-netdev``
316 - Accept offline secondaries while shutting down instance disks
317 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
318 - Added support for multiple LVM volume groups
319 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
321 - Added commands to list available fields:
323 - ``gnt-node list-fields``
324 - ``gnt-group list-fields``
325 - ``gnt-instance list-fields``
327 - Updated documentation and man pages
332 - Moved ``rapi_users`` file into separate directory, now named
333 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
335 - Added new tool for running commands on many machines,
336 ``tools/ganeti-listrunner``
337 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
338 improving the ``gnt-instance console`` output
339 - Allowed customisation of disk index separator at ``configure`` time
340 - Export node group allocation policy to :doc:`iallocator <iallocator>`
341 - Added support for non-partitioned md disks in ``lvmstrap``
342 - Added script to gracefully power off KVM instances
343 - Split ``utils`` module into smaller parts
344 - Changed query operations to return more detailed information, e.g.
345 whether an information is unavailable due to an offline node. To use
346 this new functionality, the LUXI call ``Query`` must be used. Field
347 information is now stored by the master daemon and can be retrieved
348 using ``QueryFields``. Instances, nodes and groups can also be queried
349 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
350 exposed via RAPI). The following commands make use of this
351 infrastructure change:
354 - ``gnt-group list-fields``
356 - ``gnt-node list-fields``
357 - ``gnt-instance list``
358 - ``gnt-instance list-fields``
359 - ``gnt-debug locks``
364 - New RAPI resources (see :doc:`rapi`):
368 - ``/2/groups/[group_name]``
369 - ``/2/groups/[group_name]/assign-nodes``
370 - ``/2/groups/[group_name]/modify``
371 - ``/2/groups/[group_name]/rename``
372 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
376 - Implemented ``no_install`` for instance creation
377 - Implemented OS parameters for instance reinstallation, allowing
378 use of special settings on reinstallation (e.g. for preserving data)
383 - Added IPv6 support in import/export
384 - Pause DRBD synchronization while wiping disks on instance creation
385 - Updated unittests and QA scripts
386 - Improved network parameters passed to KVM
387 - Converted man pages from docbook to reStructuredText
393 *(Released Mon, 20 Dec 2010)*
395 Released version 2.3.1~rc1 without any changes.
401 *(Released Wed, 1 Dec 2010)*
403 - impexpd: Disable OpenSSL compression in socat if possible (backport
404 from master, commit e90739d625b, see :doc:`installation guide
405 <install-quick>` for details)
406 - Changed unittest coverage report to exclude test scripts
407 - Added script to check version format
413 *(Released Wed, 1 Dec 2010)*
415 Released version 2.3.0~rc1 without any changes.
421 *(Released Fri, 19 Nov 2010)*
423 A number of bugfixes and documentation updates:
425 - Update ganeti-os-interface documentation
426 - Fixed a bug related to duplicate MACs or similar items which should be
428 - Fix breakage in OS state modify
429 - Reinstall instance: disallow offline secondaries (fixes bug related to
430 OS changing but reinstall failing)
431 - plus all the other fixes between 2.2.1 and 2.2.2
437 *(Released Tue, 2 Nov 2010)*
439 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
440 - Fixed master failover race with watcher
441 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
443 - Accept previously stopped instance for export with instance removal
444 - Simplify and extend the environment variables for instance OS scripts
445 - Added new node flags, ``master_capable`` and ``vm_capable``
446 - Added optional instance disk wiping prior during allocation. This is a
447 cluster-wide option and can be set/modified using
448 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
449 - Added IPv6 support, see :doc:`design document <design-2.3>` and
451 - Added a new watcher option (``--ignore-pause``)
452 - Added option to ignore offline node on instance start/stop
453 (``--ignore-offline``)
454 - Allow overriding OS parameters with ``gnt-instance reinstall``
455 - Added ability to change node's secondary IP address using ``gnt-node
457 - Implemented privilege separation for all daemons except
458 ``ganeti-noded``, see ``configure`` options
459 - Complain if an instance's disk is marked faulty in ``gnt-cluster
461 - Implemented job priorities (see ``ganeti(7)`` manpage)
462 - Ignore failures while shutting down instances during failover from
464 - Exit daemon's bootstrap process only once daemon is ready
465 - Export more information via ``LUInstanceQuery``/remote API
466 - Improved documentation, QA and unittests
467 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
468 restart if the file was created or changed
469 - Added LUXI protocol version sent with each request and response,
470 allowing detection of server/client mismatches
471 - Moved the Python scripts among gnt-* and ganeti-* into modules
472 - Moved all code related to setting up SSH to an external script,
474 - Infrastructure changes for node group support in future versions
480 *(Released Fri, 19 Nov 2010)*
482 A few small bugs fixed, and some improvements to the build system:
484 - Fix documentation regarding conversion to drbd
485 - Fix validation of parameters in cluster modify (``gnt-cluster modify
487 - Fix error handling in node modify with multiple changes
488 - Allow remote imports without checked names
494 *(Released Tue, 19 Oct 2010)*
496 - Disable SSL session ID cache in RPC client
502 *(Released Thu, 14 Oct 2010)*
504 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
505 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
507 - Fix problems with interaction between Curl and Python's HTTP server,
508 resulting in increased speed in many RPC calls
509 - Improve our release script to prevent breakage with older aclocal and
516 *(Released Thu, 7 Oct 2010)*
518 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
519 value retrieved from master
520 - Added support for blacklisted or hidden OS definitions
521 - Added simple lock monitor (accessible via (``gnt-debug locks``)
522 - Added support for -mem-path in KVM hypervisor abstraction layer
523 - Allow overriding instance parameters in tool for inter-cluster
524 instance moves (``tools/move-instance``)
525 - Improved opcode summaries (e.g. in ``gnt-job list``)
526 - Improve consistency of OS listing by sorting it
527 - Documentation updates
533 *(Released Fri, 8 Oct 2010)*
535 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
541 *(Released Mon, 4 Oct 2010)*
543 - Fixed regression in ``gnt-instance rename``
549 *(Released Wed, 22 Sep 2010)*
551 - Fixed OS_VARIANT variable for OS scripts
552 - Fixed cluster tag operations via RAPI
553 - Made ``setup-ssh`` exit with non-zero code if an error occurred
554 - Disabled RAPI CA checks in watcher
560 *(Released Mon, 23 Aug 2010)*
562 - Support DRBD versions of the format "a.b.c.d"
564 - Re-introduce support for usage from multiple threads in RAPI client
565 - Instance renames and modify via RAPI
566 - Work around race condition between processing and archival in job
568 - Mark opcodes following failed one as failed, too
569 - Job field ``lock_status`` was removed due to difficulties making it
570 work with the changed job queue in Ganeti 2.2; a better way to monitor
571 locks is expected for a later 2.2.x release
572 - Fixed dry-run behaviour with many commands
573 - Support ``ssh-agent`` again when adding nodes
574 - Many additional bugfixes
580 *(Released Fri, 30 Jul 2010)*
582 Important change: the internal RPC mechanism between Ganeti nodes has
583 changed from using a home-grown http library (based on the Python base
584 libraries) to use the PycURL library. This requires that PycURL is
585 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
586 against GnuTLS by default. cURL's support for GnuTLS had known issues
587 before cURL 7.21.0 and we recommend using the latest cURL release or
588 linking against OpenSSL. Most other distributions already link PycURL
589 and cURL against OpenSSL. The command::
591 python -c 'import pycurl; print pycurl.version'
593 can be used to determine the libraries PycURL and cURL are linked
596 Other significant changes:
598 - Rewrote much of the internals of the job queue, in order to achieve
599 better parallelism; this decouples job query operations from the job
600 processing, and it should allow much nicer behaviour of the master
601 daemon under load, and it also has uncovered some long-standing bugs
602 related to the job serialisation (now fixed)
603 - Added a default iallocator setting to the cluster parameters,
604 eliminating the need to always pass nodes or an iallocator for
605 operations that require selection of new node(s)
606 - Added experimental support for the LXC virtualization method
607 - Added support for OS parameters, which allows the installation of
608 instances to pass parameter to OS scripts in order to customise the
610 - Added a hypervisor parameter controlling the migration type (live or
611 non-live), since hypervisors have various levels of reliability; this
612 has renamed the 'live' parameter to 'mode'
613 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
614 logical volumes, meaning that cluster verify will ignore them and not
615 flag their presence as errors
616 - The watcher will now reset the error count for failed instances after
617 8 hours, thus allowing self-healing if the problem that caused the
618 instances to be down/fail to start has cleared in the meantime
619 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
620 check for, and warn, if the drbd module parameter ``usermode_helper``
621 is not consistent with the cluster-wide setting; this is needed to
622 make diagnose easier of failed drbd creations
623 - Started adding base IPv6 support, but this is not yet
624 enabled/available for use
625 - Rename operations (cluster, instance) will now return the new name,
626 which is especially useful if a short name was passed in
627 - Added support for instance migration in RAPI
628 - Added a tool to pre-configure nodes for the SSH setup, before joining
629 them to the cluster; this will allow in the future a simplified model
630 for node joining (but not yet fully enabled in 2.2); this needs the
631 paramiko python library
632 - Fixed handling of name-resolving errors
633 - Fixed consistency of job results on the error path
634 - Fixed master-failover race condition when executed multiple times in
636 - Fixed many bugs related to the job queue (mostly introduced during the
637 2.2 development cycle, so not all are impacting 2.1)
638 - Fixed instance migration with missing disk symlinks
639 - Fixed handling of unknown jobs in ``gnt-job archive``
640 - And many other small fixes/improvements
644 - Enhanced both the unittest and the QA coverage
645 - Switched the opcode validation to a generic model, and extended the
646 validation to all opcode parameters
647 - Changed more parts of the code that write shell scripts to use the
649 - Switched the master daemon to use the asyncore library for the Luxi
656 *(Released Thu, 17 Jun 2010)*
658 - Added tool (``move-instance``) and infrastructure to move instances
659 between separate clusters (see :doc:`separate documentation
660 <move-instance>` and :doc:`design document <design-2.2>`)
661 - Added per-request RPC timeout
662 - RAPI now requires a Content-Type header for requests with a body (e.g.
663 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
664 :rfc:`2616` (HTTP/1.1), section 7.2.1)
665 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
667 - Implemented initial support for running Ganeti daemons as separate
668 users, see configure-time flags ``--with-user-prefix`` and
669 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
671 - Instances can be removed after export (``gnt-backup export
673 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
674 key (instead of 1024 bit)
675 - Added new cluster configuration file for cluster domain secret
676 - Import/export now use SSL instead of SSH
677 - Added support for showing estimated time when exporting an instance,
678 see the ``ganeti-os-interface(7)`` manpage and look for
685 *(Released Tue, 16 Nov 2010)*
687 Some more bugfixes. Unless critical bugs occur, this will be the last
690 - Fix case of MAC special-values
691 - Fix mac checker regex
692 - backend: Fix typo causing "out of range" error
693 - Add missing --units in gnt-instance list man page
699 *(Released Tue, 24 Aug 2010)*
702 - Don't ignore secondary node silently on non-mirrored disk templates
704 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
705 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
706 - Properly document the usb_mouse parameter
707 - Fix path in ganeti-rapi(8) (issue 116)
708 - Adjust error message when the ganeti user's .ssh directory is
710 - Add same-node-check when changing the disk template to drbd
716 *(Released Fri, 16 Jul 2010)*
719 - Add an option to only select some reboot types during qa/burnin.
720 (on some hypervisors consequent reboots are not supported)
721 - Fix infrequent race condition in master failover. Sometimes the old
722 master ip address would be still detected as up for a short time
723 after it was removed, causing failover to fail.
724 - Decrease mlockall warnings when the ctypes module is missing. On
725 Python 2.4 we support running even if no ctypes module is installed,
726 but we were too verbose about this issue.
727 - Fix building on old distributions, on which man doesn't have a
729 - Fix RAPI not to ignore the MAC address on instance creation
730 - Implement the old instance creation format in the RAPI client.
736 *(Released Thu, 01 Jul 2010)*
738 A small bugfix release:
739 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
740 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
741 a non-existing option
742 - Fix instance create: the --force-variant option was ignored
743 - Improve pylint 0.21 compatibility and warnings with Python 2.6
744 - Fix modify node storage with non-FQDN arguments
745 - Fix RAPI client to authenticate under Python 2.6 when used
746 for more than 5 requests needing authentication
747 - Fix gnt-instance modify -t (storage) giving a wrong error message
748 when converting a non-shutdown drbd instance to plain
754 *(Released Fri, 18 Jun 2010)*
756 A small bugfix release:
758 - Fix live migration of KVM instances started with older Ganeti
759 versions which had fewer hypervisor parameters
760 - Fix gnt-instance grow-disk on down instances
761 - Fix an error-reporting bug during instance migration
762 - Better checking of the ``--net`` and ``--disk`` values, to avoid
763 silently ignoring broken ones
764 - Fix an RPC error reporting bug affecting, for example, RAPI client
766 - Fix bug triggered by different API version os-es on different nodes
767 - Fix a bug in instance startup with custom hvparams: OS level
768 parameters would fail to be applied.
769 - Fix the RAPI client under Python 2.6 (but more work is needed to
770 make it work completely well with OpenSSL)
771 - Fix handling of errors when resolving names from DNS
777 *(Released Thu, 3 Jun 2010)*
779 A medium sized development cycle. Some new features, and some
780 fixes/small improvements/cleanups.
785 The node deamon now tries to mlock itself into memory, unless the
786 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
787 its logs, and falls back to console logging. This allows emergency
788 features such as ``gnt-node powercycle`` to work even in the event of a
789 broken node disk (tested offlining the disk hosting the node's
790 filesystem and dropping its memory caches; don't try this at home)
792 KVM: add vhost-net acceleration support. It can be tested with a new
793 enough version of the kernel and of qemu-kvm.
795 KVM: Add instance chrooting feature. If you use privilege dropping for
796 your VMs you can also now force them to chroot to an empty directory,
797 before starting the emulated guest.
799 KVM: Add maximum migration bandwith and maximum downtime tweaking
800 support (requires a new-enough version of qemu-kvm).
802 Cluster verify will now warn if the master node doesn't have the master
805 Add a new (incompatible) instance creation request format to RAPI which
806 supports all parameters (previously only a subset was supported, and it
807 wasn't possible to extend the old format to accomodate all the new
808 features. The old format is still supported, and a client can check for
809 this feature, before using it, by checking for its presence in the
810 ``features`` RAPI resource.
812 Now with ancient latin support. Try it passing the ``--roman`` option to
813 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
814 (requires the python-roman module to be installed, in order to work).
819 As usual many internal code refactorings, documentation updates, and
822 - Lots of improvements and cleanups to the experimental Remote API
823 (RAPI) client library.
824 - A new unit test suite for the core daemon libraries.
825 - A fix to creating missing directories makes sure the umask is not
826 applied anymore. This enforces the same directory permissions
828 - Better handling terminating daemons with ctrl+c (used when running
829 them in debugging mode).
830 - Fix a race condition in live migrating a KVM instance, when stat()
831 on the old proc status file returned EINVAL, which is an unexpected
833 - Fixed manpage checking with newer man and utf-8 charachters. But now
834 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
840 *(Released Fri, 7 May 2010)*
842 Fix a bug which prevented untagged KVM instances from starting.
848 *(Released Fri, 7 May 2010)*
850 Another release with a long development cycle, during which many
851 different features were added.
856 The KVM hypervisor now can run the individual instances as non-root, to
857 reduce the impact of a VM being hijacked due to bugs in the
858 hypervisor. It is possible to run all instances as a single (non-root)
859 user, to manually specify a user for each instance, or to dynamically
860 allocate a user out of a cluster-wide pool to each instance, with the
861 guarantee that no two instances will run under the same user ID on any
864 An experimental RAPI client library, that can be used standalone
865 (without the other Ganeti libraries), is provided in the source tree as
866 ``lib/rapi/client.py``. Note this client might change its interface in
867 the future, as we iterate on its capabilities.
869 A new command, ``gnt-cluster renew-crypto`` has been added to easily
870 replace the cluster's certificates and crypto keys. This might help in
871 case they have been compromised, or have simply expired.
873 A new disk option for instance creation has been added that allows one
874 to "adopt" currently existing logical volumes, with data
875 preservation. This should allow easier migration to Ganeti from
876 unmanaged (or managed via other software) instances.
878 Another disk improvement is the possibility to convert between redundant
879 (DRBD) and plain (LVM) disk configuration for an instance. This should
880 allow better scalability (starting with one node and growing the
881 cluster, or shrinking a two-node cluster to one node).
883 A new feature that could help with automated node failovers has been
884 implemented: if a node sees itself as offline (by querying the master
885 candidates), it will try to shutdown (hard) all instances and any active
886 DRBD devices. This reduces the risk of duplicate instances if an
887 external script automatically failovers the instances on such nodes. To
888 enable this, the cluster parameter ``maintain_node_health`` should be
889 enabled; in the future this option (per the name) will enable other
890 automatic maintenance features.
892 Instance export/import now will reuse the original instance
893 specifications for all parameters; that means exporting an instance,
894 deleting it and the importing it back should give an almost identical
895 instance. Note that the default import behaviour has changed from
896 before, where it created only one NIC; now it recreates the original
899 Cluster verify has added a few new checks: SSL certificates validity,
900 /etc/hosts consistency across the cluster, etc.
905 As usual, many internal changes were done, documentation fixes,
908 - Fixed cluster initialization with disabled cluster storage (regression
910 - File-based storage supports growing the disks
911 - Fixed behaviour of node role changes
912 - Fixed cluster verify for some corner cases, plus a general rewrite of
913 cluster verify to allow future extension with more checks
914 - Fixed log spamming by watcher and node daemon (regression introduced
916 - Fixed possible validation issues when changing the list of enabled
918 - Fixed cleanup of /etc/hosts during node removal
919 - Fixed RAPI response for invalid methods
920 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
921 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
922 from ide disks, etc.)
923 - Allow OS changes without re-installation (to record a changed OS
924 outside of Ganeti, or to allow OS renames)
925 - Allow instance creation without OS installation (useful for example if
926 the OS will be installed manually, or restored from a backup not in
928 - Implemented option to make cluster ``copyfile`` use the replication
930 - Added list of enabled hypervisors to ssconf (possibly useful for
932 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
934 - A partial form of node re-IP is possible via node readd, which now
935 allows changed node primary IP
936 - Command line utilities now show an informational message if the job is
938 - The logs of the master daemon now show the PID/UID/GID of the
945 *(Released Fri, 12 Mar 2010)*
947 During the 2.1.0 long release candidate cycle, a lot of improvements and
948 changes have accumulated with were released later as 2.1.1.
953 The node evacuate command (``gnt-node evacuate``) was significantly
954 rewritten, and as such the IAllocator protocol was changed - a new
955 request type has been added. This unfortunate change during a stable
956 series is designed to improve performance of node evacuations; on
957 clusters with more than about five nodes and which are well-balanced,
958 evacuation should proceed in parallel for all instances of the node
959 being evacuated. As such, any existing IAllocator scripts need to be
960 updated, otherwise the above command will fail due to the unknown
961 request. The provided "dumb" allocator has not been updated; but the
962 ganeti-htools package supports the new protocol since version 0.2.4.
964 Another important change is increased validation of node and instance
965 names. This might create problems in special cases, if invalid host
966 names are being used.
968 Also, a new layer of hypervisor parameters has been added, that sits at
969 OS level between the cluster defaults and the instance ones. This allows
970 customisation of virtualization parameters depending on the installed
971 OS. For example instances with OS 'X' may have a different KVM kernel
972 (or any other parameter) than the cluster defaults. This is intended to
973 help managing a multiple OSes on the same cluster, without manual
974 modification of each instance's parameters.
976 A tool for merging clusters, ``cluster-merge``, has been added in the
982 - Improved the int/float conversions that should make the code more
983 robust in face of errors from the node daemons
984 - Fixed the remove node code in case of internal configuration errors
985 - Fixed the node daemon behaviour in face of inconsistent queue
986 directory (e.g. read-only file-system where we can't open the files
988 - Fixed the behaviour of gnt-node modify for master candidate demotion;
989 now it either aborts cleanly or, if given the new "auto_promote"
990 parameter, will automatically promote other nodes as needed
991 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
992 completely prevent Ganeti from working
993 - Fixed bug for instance export when not all disks were successfully
995 - Fixed behaviour of node add when the new node is slow in starting up
997 - Fixed handling of signals in the LUXI client, which should improve
998 behaviour of command-line scripts
999 - Added checks for invalid node/instance names in the configuration (now
1000 flagged during cluster verify)
1001 - Fixed watcher behaviour for disk activation errors
1002 - Fixed two potentially endless loops in http library, which led to the
1003 RAPI daemon hanging and consuming 100% CPU in some cases
1004 - Fixed bug in RAPI daemon related to hashed passwords
1005 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1007 - Enhanced compatibility with non-Debian OSes, but not using absolute
1008 path in some commands and allowing customisation of the ssh
1009 configuration directory
1010 - Fixed possible future issue with new Python versions by abiding to the
1011 proper use of ``__slots__`` attribute on classes
1012 - Added checks that should prevent directory traversal attacks
1013 - Many documentation fixes based on feedback from users
1018 - Added an "early_release" more for instance replace disks and node
1019 evacuate, where we release locks earlier and thus allow higher
1020 parallelism within the cluster
1021 - Added watcher hooks, intended to allow the watcher to restart other
1022 daemons (e.g. from the ganeti-nbma project), but they can be used of
1023 course for any other purpose
1024 - Added a compile-time disable for DRBD barriers, to increase
1025 performance if the administrator trusts the power supply or the
1026 storage system to not lose writes
1027 - Added the option of using syslog for logging instead of, or in
1028 addition to, Ganeti's own log files
1029 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1030 can indeed boot from a paravirtual NIC
1031 - Added a generic debug level for many operations; while this is not
1032 used widely yet, it allows one to pass the debug value all the way to
1034 - Enhanced the hooks environment for instance moves (failovers,
1035 migrations) where the primary/secondary nodes changed during the
1036 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1037 - Enhanced data validations for many user-supplied values; one important
1038 item is the restrictions imposed on instance and node names, which
1039 might reject some (invalid) host names
1040 - Add a configure-time option to disable file-based storage, if it's not
1041 needed; this allows greater security separation between the master
1042 node and the other nodes from the point of view of the inter-node RPC
1044 - Added user notification in interactive tools if job is waiting in the
1045 job queue or trying to acquire locks
1046 - Added log messages when a job is waiting for locks
1047 - Added filtering by node tags in instance operations which admit
1048 multiple instances (start, stop, reboot, reinstall)
1049 - Added a new tool for cluster mergers, ``cluster-merge``
1050 - Parameters from command line which are of the form ``a=b,c=d`` can now
1051 use backslash escapes to pass in values which contain commas,
1052 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1054 - For KVM, the instance name is the first parameter passed to KVM, so
1055 that it's more visible in the process list
1061 *(Released Tue, 2 Mar 2010)*
1063 Ganeti 2.1 brings many improvements with it. Major changes:
1065 - Added infrastructure to ease automated disk repairs
1066 - Added new daemon to export configuration data in a cheaper way than
1067 using the remote API
1068 - Instance NICs can now be routed instead of being associated with a
1070 - Improved job locking logic to reduce impact of jobs acquiring multiple
1071 locks waiting for other long-running jobs
1073 In-depth implementation details can be found in the Ganeti 2.1 design
1079 - Added chroot hypervisor
1080 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1082 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1083 ``bootloader_path`` and ``bootloader_args``)
1084 - Added the ``use_localtime`` option for the xen-hvm and kvm
1085 hypervisors, and the default value for this has changed to false (in
1086 2.0 xen-hvm always enabled it)
1087 - Added luxi call to submit multiple jobs in one go
1088 - Added cluster initialization option to not modify ``/etc/hosts``
1090 - Added network interface parameters
1091 - Added dry run mode to some LUs
1092 - Added RAPI resources:
1094 - ``/2/instances/[instance_name]/info``
1095 - ``/2/instances/[instance_name]/replace-disks``
1096 - ``/2/nodes/[node_name]/evacuate``
1097 - ``/2/nodes/[node_name]/migrate``
1098 - ``/2/nodes/[node_name]/role``
1099 - ``/2/nodes/[node_name]/storage``
1100 - ``/2/nodes/[node_name]/storage/modify``
1101 - ``/2/nodes/[node_name]/storage/repair``
1103 - Added OpCodes to evacuate or migrate all instances on a node
1104 - Added new command to list storage elements on nodes (``gnt-node
1105 list-storage``) and modify them (``gnt-node modify-storage``)
1106 - Added new ssconf files with master candidate IP address
1107 (``ssconf_master_candidates_ips``), node primary IP address
1108 (``ssconf_node_primary_ips``) and node secondary IP address
1109 (``ssconf_node_secondary_ips``)
1110 - Added ``ganeti-confd`` and a client library to query the Ganeti
1111 configuration via UDP
1112 - Added ability to run hooks after cluster initialization and before
1114 - Added automatic mode for disk replace (``gnt-instance replace-disks
1116 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1117 after catastrophic data-loss
1118 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1120 - Added ``gnt-instance move`` command to move instances
1121 - Added ``gnt-cluster watcher`` command to control watcher
1122 - Added ``gnt-node powercycle`` command to powercycle nodes
1123 - Added new job status field ``lock_status``
1124 - Added parseable error codes to cluster verification (``gnt-cluster
1125 verify --error-codes``) and made output less verbose (use
1126 ``--verbose`` to restore previous behaviour)
1127 - Added UUIDs to the main config entities (cluster, nodes, instances)
1128 - Added support for OS variants
1129 - Added support for hashed passwords in the Ganeti remote API users file
1131 - Added option to specify maximum timeout on instance shutdown
1132 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1133 - Added new helper script to start and stop Ganeti daemons
1134 (``daemon-util``), with the intent to reduce the work necessary to
1135 adjust Ganeti for non-Debian distributions and to start/stop daemons
1137 - Added more unittests
1138 - Fixed critical bug in ganeti-masterd startup
1139 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1140 now customisable at the cluster level for both the KVM and Xen
1141 hypervisors using the new ``migration_port`` parameter
1142 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1143 reinstalling an instance
1144 - Allowed ``@`` in tag names
1145 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1146 - Many documentation updates
1147 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1148 - ``gnt-instance reinstall`` can now reinstall multiple instances
1149 - Updated many command line parameters
1150 - Introduced new OS API version 15
1151 - No longer support a default hypervisor
1152 - Treat virtual LVs as inexistent
1153 - Improved job locking logic to reduce lock contention
1154 - Match instance and node names case insensitively
1155 - Reimplemented bash completion script to be more complete
1162 *(Released Thu, 4 Feb 2010)*
1164 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1165 - Fix a string formatting bug
1166 - Improve safety of the code in some error paths
1167 - Improve data validation in the master of values returned from nodes
1173 *(Released Thu, 17 Dec 2009)*
1175 - Fix security issue due to missing validation of iallocator names; this
1176 allows local and remote execution of arbitrary executables
1177 - Fix failure of gnt-node list during instance removal
1178 - Ship the RAPI documentation in the archive
1184 *(Released Wed, 30 Sep 2009)*
1186 - Fixed many wrong messages
1187 - Fixed a few bugs related to the locking library
1188 - Fixed MAC checking at instance creation time
1189 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1190 - Fixed a few issues related to signal handling in both daemons and
1192 - Fixed the example startup script provided
1193 - Fixed insserv dependencies in the example startup script (patch from
1195 - Fixed handling of drained nodes in the iallocator framework
1196 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1198 - Fixed error related to invalid job IDs in job polling
1199 - Fixed job/opcode persistence on unclean master shutdown
1200 - Fixed handling of partial job processing after unclean master
1202 - Fixed error reporting from LUs, previously all errors were converted
1203 into execution errors
1204 - Fixed error reporting from burnin
1205 - Decreased significantly the memory usage of the job queue
1206 - Optimised slightly multi-job submission
1207 - Optimised slightly opcode loading
1208 - Backported the multi-job submit framework from the development
1209 branch; multi-instance start and stop should be faster
1210 - Added script to clean archived jobs after 21 days; this will reduce
1211 the size of the queue directory
1212 - Added some extra checks in disk size tracking
1213 - Added an example ethers hook script
1214 - Added a cluster parameter that prevents Ganeti from modifying of
1216 - Added more node information to RAPI responses
1217 - Added a ``gnt-job watch`` command that allows following the ouput of a
1219 - Added a bind-address option to ganeti-rapi
1220 - Added more checks to the configuration verify
1221 - Enhanced the burnin script such that some operations can be retried
1223 - Converted instance reinstall to multi-instance model
1229 *(Released Fri, 7 Aug 2009)*
1231 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1232 to allow using the pre-2.0.2 behaviour in activation, if any existing
1233 instances have mismatched disk sizes in the configuration
1234 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1235 any configuration mismatches for disk sizes
1236 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1237 failover to work on two-node clusters
1238 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1240 - Fixed detection of OS script errors in ``gnt-backup export``
1241 - Fixed exit code of ``gnt-backup export``
1247 *(Released Fri, 17 Jul 2009)*
1249 - Added experimental support for stripped logical volumes; this should
1250 enhance performance but comes with a higher complexity in the block
1251 device handling; stripping is only enabled when passing
1252 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1253 affected even in the non-stripped mode
1254 - Improved resiliency against transient failures at the end of DRBD
1255 resyncs, and in general of DRBD resync checks
1256 - Fixed a couple of issues with exports and snapshot errors
1257 - Fixed a couple of issues in instance listing
1258 - Added display of the disk size in ``gnt-instance info``
1259 - Fixed checking for valid OSes in instance creation
1260 - Fixed handling of the "vcpus" parameter in instance listing and in
1261 general of invalid parameters
1262 - Fixed http server library, and thus RAPI, to handle invalid
1263 username/password combinations correctly; this means that now they
1264 report unauthorized for queries too, not only for modifications,
1265 allowing earlier detect of configuration problems
1266 - Added a new "role" node list field, equivalent to the master/master
1267 candidate/drained/offline flags combinations
1268 - Fixed cluster modify and changes of candidate pool size
1269 - Fixed cluster verify error messages for wrong files on regular nodes
1270 - Fixed a couple of issues with node demotion from master candidate role
1271 - Fixed node readd issues
1272 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1273 - Added a new ``--no-voting`` option for masterfailover to fix failover
1274 on two-nodes clusters when the former master node is unreachable
1275 - Added instance reinstall over RAPI
1281 *(Released Tue, 16 Jun 2009)*
1283 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1284 allow re-adding the start in single-user option (regression from 1.2)
1285 - the watcher writes the instance status to a file, to allow monitoring
1286 to report the instance status (from the master) based on cached
1287 results of the watcher's queries; while this can get stale if the
1288 watcher is being locked due to other work on the cluster, this is
1289 still an improvement
1290 - the watcher now also restarts the node daemon and the rapi daemon if
1292 - fixed the watcher to handle full and drained queue cases
1293 - hooks export more instance data in the environment, which helps if
1294 hook scripts need to take action based on the instance's properties
1295 (no longer need to query back into ganeti)
1296 - instance failovers when the instance is stopped do not check for free
1297 RAM, so that failing over a stopped instance is possible in low memory
1299 - rapi uses queries for tags instead of jobs (for less job traffic), and
1300 for cluster tags it won't talk to masterd at all but read them from
1302 - a couple of error handling fixes in RAPI
1303 - drbd handling: improved the error handling of inconsistent disks after
1304 resync to reduce the frequency of "there are some degraded disks for
1305 this instance" messages
1306 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1307 error handling path called a wrong function name)
1313 *(Released Wed, 27 May 2009)*
1315 - no changes from rc5
1318 Version 2.0 release candidate 5
1319 -------------------------------
1321 *(Released Wed, 20 May 2009)*
1323 - fix a couple of bugs (validation, argument checks)
1324 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1325 - some small improvements to RAPI and IAllocator
1326 - make watcher automatically start the master daemon if down
1329 Version 2.0 release candidate 4
1330 -------------------------------
1332 *(Released Mon, 27 Apr 2009)*
1334 - change the OS list to not require locks; this helps with big clusters
1335 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1336 volume group is broken
1337 - ``gnt-instance info``, without any arguments, doesn't run for all
1338 instances anymore; either pass ``--all`` or pass the desired
1339 instances; this helps against mistakes on big clusters where listing
1340 the information for all instances takes a long time
1341 - miscellaneous doc and man pages fixes
1344 Version 2.0 release candidate 3
1345 -------------------------------
1347 *(Released Wed, 8 Apr 2009)*
1349 - Change the internal locking model of some ``gnt-node`` commands, in
1350 order to reduce contention (and blocking of master daemon) when
1351 batching many creation/reinstall jobs
1352 - Fixes to Xen soft reboot
1353 - No longer build documentation at build time, instead distribute it in
1354 the archive, in order to reduce the need for the whole docbook/rst
1358 Version 2.0 release candidate 2
1359 -------------------------------
1361 *(Released Fri, 27 Mar 2009)*
1363 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1364 - Fix watcher startup sequence, improves the behaviour of busy clusters
1365 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1366 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1367 bind address and other places
1368 - Some documentation fixes and updates
1371 Version 2.0 release candidate 1
1372 -------------------------------
1374 *(Released Mon, 2 Mar 2009)*
1376 - More documentation updates, now all docs should be more-or-less
1378 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1380 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1382 - Fix an issue related to $libdir/run/ganeti and cluster creation
1388 *(Released Thu, 19 Feb 2009)*
1390 - Xen PVM and KVM have switched the default value for the instance root
1391 disk to the first partition on the first drive, instead of the whole
1392 drive; this means that the OS installation scripts must be changed
1394 - Man pages have been updated
1395 - RAPI has been switched by default to HTTPS, and the exported functions
1396 should all work correctly
1397 - RAPI v1 has been removed
1398 - Many improvements to the KVM hypervisor
1399 - Block device errors are now better reported
1400 - Many other bugfixes and small improvements
1406 *(Released Mon, 26 Jan 2009)*
1408 - Version 2 is a general rewrite of the code and therefore the
1409 differences are too many to list, see the design document for 2.0 in
1410 the ``doc/`` subdirectory for more details
1411 - In this beta version there is not yet a migration path from 1.2 (there
1412 will be one in the final 2.0 release)
1413 - A few significant changes are:
1415 - all commands are executed by a daemon (``ganeti-masterd``) and the
1416 various ``gnt-*`` commands are just front-ends to it
1417 - all the commands are entered into, and executed from a job queue,
1418 see the ``gnt-job(8)`` manpage
1419 - the RAPI daemon supports read-write operations, secured by basic
1420 HTTP authentication on top of HTTPS
1421 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1422 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1423 to migrate to DRBD 8 first while still running Ganeti 1.2)
1424 - DRBD devices are using statically allocated minor numbers, which
1425 will be assigned to existing instances during the migration process
1426 - there is support for both Xen PVM and Xen HVM instances running on
1428 - KVM virtualization is supported too
1429 - file-based storage has been implemented, which means that it is
1430 possible to run the cluster without LVM and DRBD storage, for
1431 example using a shared filesystem exported from shared storage (and
1432 still have live migration)
1438 *(Released Tue, 13 Jan 2009)*
1440 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1441 - Reuse the old instance mac address by default on instance import, if
1442 the instance name is the same.
1443 - Handle situations in which the node info rpc returns incomplete
1445 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1446 - Improved version of batcher:
1448 - state file support
1449 - instance mac address support
1450 - support for HVM clusters/instances
1452 - Add an option to show the number of cpu sockets and nodes in
1454 - Support OSes that handle more than one version of the OS api (but do
1455 not change the current API in any other way)
1456 - Fix ``gnt-node migrate``
1457 - ``gnt-debug`` man page
1458 - Fixes various more typos and small issues
1459 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1465 *(Released Wed, 24 Sep 2008)*
1467 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1468 type of disk exported to fully virtualized instances.
1469 - provide access to the serial console of HVM instances
1470 - instance auto_balance flag, set by default. If turned off it will
1471 avoid warnings on cluster verify if there is not enough memory to fail
1472 over an instance. in the future it will prevent automatically failing
1473 it over when we will support that.
1474 - batcher tool for instance creation, see ``tools/README.batcher``
1475 - ``gnt-instance reinstall --select-os`` to interactively select a new
1476 operating system when reinstalling an instance.
1477 - when changing the memory amount on instance modify a check has been
1478 added that the instance will be able to start. also warnings are
1479 emitted if the instance will not be able to fail over, if auto_balance
1481 - documentation fixes
1482 - sync fields between ``gnt-instance list/modify/add/import``
1483 - fix a race condition in drbd when the sync speed was set after giving
1484 the device a remote peer.
1490 *(Released Tue, 22 Jul 2008)*
1492 - note: the allowed size and number of tags per object were reduced
1493 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1494 - fixed twisted 8.x compatibility
1495 - fixed ``gnt-instance replace-disks`` with iallocator
1496 - add TCP keepalives on twisted connections to detect restarted nodes
1497 - disk increase support, see ``gnt-instance grow-disk``
1498 - implement bulk node/instance query for RAPI
1499 - add tags in node/instance listing (optional)
1500 - experimental migration (and live migration) support, read the man page
1501 for ``gnt-instance migrate``
1502 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1503 has some small improvements in handling its state file
1509 *(Released Fri, 13 Jun 2008)*
1511 - Experimental readonly, REST-based remote API implementation;
1512 automatically started on master node, TCP port 5080, if enabled by
1513 ``--enable-rapi`` parameter to configure script.
1514 - Instance allocator support. Add and import instance accept a
1515 ``--iallocator`` parameter, and call that instance allocator to decide
1516 which node to use for the instance. The iallocator document describes
1517 what's expected from an allocator script.
1518 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1519 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1520 node is lost there is still enough memory to fail over the instances
1522 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1523 ``gnt-cluster verify``, to check for site-specific compliance. All the
1524 hooks will run, and their output, if any, will be displayed. Any
1525 failing hook will make the verification return an error value.
1526 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1527 primary and secondary interfaces
1528 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1529 that is still declared as part of the cluster and has failed.
1530 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1531 specifying output fields, that just adds the chosen fields to the
1533 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1534 export from the filesystem.
1535 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1536 have been added. Using them you can enable/disable acpi and pae
1537 support, and specify a path for a cd image to be exported to the
1538 instance. These parameters as the name suggest only work on HVM
1540 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1541 PAE support will be set to the previously hardcoded values, but the
1542 (previously hardcoded) path to the CDROM ISO image will be unset and
1543 if required, needs to be set manually with ``gnt-instance modify``
1545 - The address to which an instance's VNC console is bound is now
1546 selectable per-instance, rather than being cluster wide. Of course
1547 this only applies to instances controlled via VNC, so currently just
1548 applies to HVM clusters.
1554 *(Released Mon, 18 Feb 2008)*
1556 - more tweaks to the disk activation code (especially helpful for DRBD)
1557 - change the default ``gnt-instance list`` output format, now there is
1558 one combined status field (see the manpage for the exact values this
1560 - some more fixes for the mac export to hooks change
1561 - make Ganeti not break with DRBD 8.2.x (which changed the version
1562 format in ``/proc/drbd``) (issue 24)
1563 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1564 template, allowing migration from DRBD0.7+MD to DRBD8
1570 *(Released Wed, 30 Jan 2008)*
1572 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1574 - add command aliases infrastructure and a few aliases
1575 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1576 man pages and the ``--help`` option of ``gnt-node
1577 list``/``gnt-instance list``
1578 - fix ``gnt-backup list`` with down nodes (issue 21)
1579 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1580 - fix the dist archive and add a check for including svn/git files in
1582 - some developer-related changes: improve the burnin and the QA suite,
1583 add an upload script for testing during development
1589 *(Released Wed, 16 Jan 2008)*
1591 - experimental HVM support, read the install document, section
1592 "Initializing the cluster"
1593 - allow for the PVM hypervisor per-instance kernel and initrd paths
1594 - add a new command ``gnt-cluster verify-disks`` which uses a new
1595 algorithm to improve the reconnection of the DRBD pairs if the device
1596 on the secondary node has gone away
1597 - make logical volume code auto-activate LVs at disk activation time
1598 - slightly improve the speed of activating disks
1599 - allow specification of the MAC address at instance creation time, and
1600 changing it later via ``gnt-instance modify``
1601 - fix handling of external commands that generate lots of output on
1603 - update documentation with regard to minimum version of DRBD8 supported
1609 *(Released Tue, 4 Dec 2007)*
1611 - Log the ``xm create`` output to the node daemon log on failure (to
1612 help diagnosing the error)
1613 - In debug mode, log all external commands output if failed to the logs
1614 - Change parsing of lvm commands to ignore stderr
1620 *(Released Wed, 28 Nov 2007)*
1622 - Another round of updates to the DRBD 8 code to deal with more failures
1623 in the replace secondary node operation
1624 - Some more logging of failures in disk operations (lvm, drbd)
1625 - A few documentation updates
1632 *(Released Tue, 13 Nov 2007)*
1634 - Change configuration file format from Python's Pickle to JSON.
1635 Upgrading is possible using the cfgupgrade utility.
1636 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1637 faster replace disks and is more stable (DRBD 8 has many improvements
1638 compared to DRBD 0.7)
1639 - Added command line tags support (see man pages for ``gnt-instance``,
1640 ``gnt-node``, ``gnt-cluster``)
1641 - Added instance rename support
1642 - Added multi-instance startup/shutdown
1643 - Added cluster rename support
1644 - Added ``gnt-node evacuate`` to simplify some node operations
1645 - Added instance reboot operation that can speedup reboot as compared to
1647 - Soften the requirement that hostnames are in FQDN format
1648 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1650 - Removed dependency on debian's patched fping that uses the
1651 non-standard ``-S`` option
1652 - Now the OS definitions are searched for in multiple, configurable
1653 paths (easier for distros to package)
1654 - Some changes to the hooks infrastructure (especially the new
1655 post-configuration update hook)
1656 - Other small bugfixes
1658 .. vim: set textwidth=72 syntax=rst :