7 *(Released Tue, 23 Aug 2011)*
11 - Fixed documentation for importing with ``--src-dir`` option
12 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
13 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
19 *(Released Fri, 5 Aug 2011)*
21 Many bug-fixes and a few small features:
23 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
24 issues when you tried to add an instance with two MAC addresses in one
26 - KVM: fixed per-instance stored UID value
27 - KVM: configure bridged NICs at migration start
28 - KVM: Fix a bug where instance will not start with never KVM versions
30 - Added OS search path to ``gnt-cluster info``
31 - Fixed an issue with ``file_storage_dir`` where you were forced to
32 provide an absolute path, but the documentation states it is a
33 relative path, the documentation was right
34 - Added a new parameter to instance stop/start called ``--no-remember``
35 that will make the state change to not be remembered
36 - Implemented ``no_remember`` at RAPI level
37 - Improved the documentation
38 - Node evacuation: don't call IAllocator if node is already empty
39 - Fixed bug in DRBD8 replace disks on current nodes
40 - Fixed bug in recreate-disks for DRBD instances
41 - Moved assertion checking locks in ``gnt-instance replace-disks``
42 causing it to abort with not owning the right locks for some situation
43 - Job queue: Fixed potential race condition when cancelling queued jobs
44 - Fixed off-by-one bug in job serial generation
45 - ``gnt-node volumes``: Fix instance names
46 - Fixed aliases in bash completion
47 - Fixed a bug in reopening log files after being sent a SIGHUP
48 - Added a flag to burnin to allow specifying VCPU count
49 - Bugfixes to non-root Ganeti configuration
55 *(Released Thu, 12 May 2011)*
57 Many bug-fixes and a few new small features:
59 - Fixed a bug related to log opening failures
60 - Fixed a bug in instance listing with orphan instances
61 - Fixed a bug which prevented resetting the cluster-level node parameter
62 ``oob_program`` to the default
63 - Many fixes related to the ``cluster-merge`` tool
64 - Fixed a race condition in the lock monitor, which caused failures
65 during (at least) creation of many instances in parallel
66 - Improved output for gnt-job info
67 - Removed the quiet flag on some ssh calls which prevented debugging
69 - Improved the N+1 failure messages in cluster verify by actually
70 showing the memory values (needed and available)
71 - Increased lock attempt timeouts so that when executing long operations
72 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
73 too early and thus prevent the use of the 'fair' mechanism
74 - Changed instance query data (``gnt-instance info``) to not acquire
75 locks unless needed, thus allowing its use on locked instance if only
76 static information is asked for
77 - Improved behaviour with filesystems that do not support rename on an
79 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
80 kept locks on all nodes during the wipe, which is unneeded
81 - Fixed ``gnt-watcher`` handling of errors during hooks execution
82 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
83 10GiB) which caused the wipe to fail right at the end in some cases
84 - Fixed master IP activation when doing master failover with no-voting
85 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
86 the master node itself
87 - Fixed potential data-loss in under disk full conditions, where Ganeti
88 wouldn't check correctly the return code and would consider
89 partially-written files 'correct'
90 - Fixed bug related to multiple VGs and DRBD disk replacing
91 - Added new disk parameter ``metavg`` that allows placement of the meta
92 device for DRBD in a different volume group
93 - Fixed error handling in the node daemon when the system libc doesn't
94 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
95 - Fixed lock release during replace-disks, which kept cluster-wide locks
96 when doing disk replaces with an iallocator script
97 - Added check for missing bridges in cluster verify
98 - Handle EPIPE errors while writing to the terminal better, so that
99 piping the output to e.g. ``less`` doesn't cause a backtrace
100 - Fixed rare case where a ^C during Luxi calls could have been
101 interpreted as server errors, instead of simply terminating
102 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
104 - Added a few more parameters to the KVM hypervisor, allowing a second
105 CDROM, custom disk type for CDROMs and a floppy image
106 - Removed redundant message in instance rename when the name is given
108 - Added option to ``gnt-instance recreate-disks`` to allow creating the
109 disks on new nodes, allowing recreation when the original instance
110 nodes are completely gone
111 - Added option when converting disk templates to DRBD to skip waiting
112 for the resync, in order to make the instance available sooner
113 - Added two new variables to the OS scripts environment (containing the
115 - Made the root_path and optional parameter for the xen-pvm hypervisor,
116 to allow use of ``pvgrub`` as bootloader
117 - Changed the instance memory modifications to only check out-of-memory
118 conditions on memory increases, and turned the secondary node warnings
119 into errors (they can still be overridden via ``--force``)
120 - Fixed the handling of a corner case when the Python installation gets
121 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
122 to execute a command that doesn't exist
123 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
124 node of the instance returned failures during instance shutdown; this
125 adds the option ``--ignore-consistency`` to gnt-instance move
127 And as usual, various improvements to the error messages, documentation
133 *(Released Wed, 09 Mar 2011)*
135 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
136 the RAPI users file if run twice (even with ``--dry-run``).
138 The release fixes that bug (nothing else changed).
144 *(Released Mon, 07 Mar 2011)*
146 Final 2.4.0 release. Just a few small fixes:
148 - Fixed RAPI node evacuate
149 - Fixed the kvm-ifup script
150 - Fixed internal error handling for special job cases
151 - Updated man page to specify the escaping feature for options
157 *(Released Mon, 28 Feb 2011)*
159 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
160 that this feature wiped the disks of the wrong instance, leading to loss
165 - Fixed title of query field containing instance name
166 - Expanded the glossary in the documentation
167 - Fixed one unittest (internal issue)
173 *(Released Mon, 21 Feb 2011)*
175 A number of bug fixes plus just a couple functionality changes.
177 On the user-visible side, the ``gnt-* list`` command output has changed
178 with respect to "special" field states. The current rc1 style of display
179 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
180 the default output mode special fields states are displayed as follows:
182 - Offline resource: ``*``
183 - Unavailable/not applicable: ``-``
184 - Data missing (RPC failure): ``?``
185 - Unknown field: ``??``
187 Another user-visible change is the addition of ``--force-join`` to
192 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
193 - Fixed regression in RAPI/instance reinstall where all parameters were
194 required (instead of optional)
195 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
196 - Fixed iallocator usage (offline nodes were not considered offline)
197 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
198 - Fixed hypervisor and OS parameter validation with respect to
200 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
202 - Fixed ``tools/listrunner`` with respect to agent-based usage
208 *(Released Fri, 4 Feb 2011)*
210 Many changes and fixes since the beta1 release. While there were some
211 internal changes, the code has been mostly stabilised for the RC
214 Note: the dumb allocator was removed in this release, as it was not kept
215 up-to-date with the IAllocator protocol changes. It is recommended to
216 use the ``hail`` command from the ganeti-htools package.
218 Note: the 2.4 and up versions of Ganeti are not compatible with the
219 0.2.x branch of ganeti-htools. You need to upgrade to
220 ganeti-htools-0.3.0 (or later).
222 Regressions fixed from 2.3
223 ~~~~~~~~~~~~~~~~~~~~~~~~~~
225 - Fixed the ``gnt-cluster verify-disks`` command
226 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
228 - Fixed disk adoption breakage
229 - Fixed wrong headers in instance listing for field aliases
234 - Fixed corner case in KVM handling of NICs
235 - Fixed many cases of wrong handling of non-vm_capable nodes
236 - Fixed a bug where a missing instance symlink was not possible to
237 recreate with any ``gnt-*`` command (now ``gnt-instance
238 activate-disks`` does it)
239 - Fixed the volume group name as reported by ``gnt-cluster
241 - Increased timeouts for the import-export code, hopefully leading to
242 fewer aborts due network or instance timeouts
243 - Fixed bug in ``gnt-node list-storage``
244 - Fixed bug where not all daemons were started on cluster
245 initialisation, but only at the first watcher run
246 - Fixed many bugs in the OOB implementation
247 - Fixed watcher behaviour in presence of instances with offline
249 - Fixed instance list output for instances running on the wrong node
250 - a few fixes to the cluster-merge tool, but it still cannot merge
251 multi-node groups (currently it is not recommended to use this tool)
257 - Improved network configuration for the KVM hypervisor
258 - Added e1000 as a supported NIC for Xen-HVM
259 - Improved the lvmstrap tool to also be able to use partitions, as
260 opposed to full disks
261 - Improved speed of disk wiping (the cluster parameter
262 ``prealloc_wipe_disks``, so that it has a low impact on the total time
263 of instance creations
264 - Added documentation for the OS parameters
265 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
266 hypervisor is not responding
267 - Added display of blacklisted and hidden OS information in
269 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
270 NIC and node parameters, which might create problems with currently
271 invalid (but undetected) configuration files, but prevents validation
272 failures when unrelated parameters are modified
273 - Changed cluster initialisation to wait for the master daemon to become
275 - Expanded the RAPI interface:
277 - Added config redistribution resource
278 - Added activation/deactivation of instance disks
279 - Added export of console information
281 - Implemented log file reopening on SIGHUP, which allows using
282 logrotate(8) for the Ganeti log files
283 - Added a basic OOB helper script as an example
289 *(Released Fri, 14 Jan 2011)*
294 - Fixed timezone issues when formatting timestamps
295 - Added support for node groups, available via ``gnt-group`` and other
297 - Added out-of-band framework and management, see :doc:`design
298 document <design-oob>`
299 - Removed support for roman numbers from ``gnt-node list`` and
300 ``gnt-instance list``.
301 - Allowed modification of master network interface via ``gnt-cluster
302 modify --master-netdev``
303 - Accept offline secondaries while shutting down instance disks
304 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
305 - Added support for multiple LVM volume groups
306 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
308 - Added commands to list available fields:
310 - ``gnt-node list-fields``
311 - ``gnt-group list-fields``
312 - ``gnt-instance list-fields``
314 - Updated documentation and man pages
319 - Moved ``rapi_users`` file into separate directory, now named
320 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
322 - Added new tool for running commands on many machines,
323 ``tools/ganeti-listrunner``
324 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
325 improving the ``gnt-instance console`` output
326 - Allowed customisation of disk index separator at ``configure`` time
327 - Export node group allocation policy to :doc:`iallocator <iallocator>`
328 - Added support for non-partitioned md disks in ``lvmstrap``
329 - Added script to gracefully power off KVM instances
330 - Split ``utils`` module into smaller parts
331 - Changed query operations to return more detailed information, e.g.
332 whether an information is unavailable due to an offline node. To use
333 this new functionality, the LUXI call ``Query`` must be used. Field
334 information is now stored by the master daemon and can be retrieved
335 using ``QueryFields``. Instances, nodes and groups can also be queried
336 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
337 exposed via RAPI). The following commands make use of this
338 infrastructure change:
341 - ``gnt-group list-fields``
343 - ``gnt-node list-fields``
344 - ``gnt-instance list``
345 - ``gnt-instance list-fields``
346 - ``gnt-debug locks``
351 - New RAPI resources (see :doc:`rapi`):
355 - ``/2/groups/[group_name]``
356 - ``/2/groups/[group_name]/assign-nodes``
357 - ``/2/groups/[group_name]/modify``
358 - ``/2/groups/[group_name]/rename``
359 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
363 - Implemented ``no_install`` for instance creation
364 - Implemented OS parameters for instance reinstallation, allowing
365 use of special settings on reinstallation (e.g. for preserving data)
370 - Added IPv6 support in import/export
371 - Pause DRBD synchronization while wiping disks on instance creation
372 - Updated unittests and QA scripts
373 - Improved network parameters passed to KVM
374 - Converted man pages from docbook to reStructuredText
380 *(Released Mon, 20 Dec 2010)*
382 Released version 2.3.1~rc1 without any changes.
388 *(Released Wed, 1 Dec 2010)*
390 - impexpd: Disable OpenSSL compression in socat if possible (backport
391 from master, commit e90739d625b, see :doc:`installation guide
392 <install-quick>` for details)
393 - Changed unittest coverage report to exclude test scripts
394 - Added script to check version format
400 *(Released Wed, 1 Dec 2010)*
402 Released version 2.3.0~rc1 without any changes.
408 *(Released Fri, 19 Nov 2010)*
410 A number of bugfixes and documentation updates:
412 - Update ganeti-os-interface documentation
413 - Fixed a bug related to duplicate MACs or similar items which should be
415 - Fix breakage in OS state modify
416 - Reinstall instance: disallow offline secondaries (fixes bug related to
417 OS changing but reinstall failing)
418 - plus all the other fixes between 2.2.1 and 2.2.2
424 *(Released Tue, 2 Nov 2010)*
426 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
427 - Fixed master failover race with watcher
428 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
430 - Accept previously stopped instance for export with instance removal
431 - Simplify and extend the environment variables for instance OS scripts
432 - Added new node flags, ``master_capable`` and ``vm_capable``
433 - Added optional instance disk wiping prior during allocation. This is a
434 cluster-wide option and can be set/modified using
435 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
436 - Added IPv6 support, see :doc:`design document <design-2.3>` and
438 - Added a new watcher option (``--ignore-pause``)
439 - Added option to ignore offline node on instance start/stop
440 (``--ignore-offline``)
441 - Allow overriding OS parameters with ``gnt-instance reinstall``
442 - Added ability to change node's secondary IP address using ``gnt-node
444 - Implemented privilege separation for all daemons except
445 ``ganeti-noded``, see ``configure`` options
446 - Complain if an instance's disk is marked faulty in ``gnt-cluster
448 - Implemented job priorities (see ``ganeti(7)`` manpage)
449 - Ignore failures while shutting down instances during failover from
451 - Exit daemon's bootstrap process only once daemon is ready
452 - Export more information via ``LUInstanceQuery``/remote API
453 - Improved documentation, QA and unittests
454 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
455 restart if the file was created or changed
456 - Added LUXI protocol version sent with each request and response,
457 allowing detection of server/client mismatches
458 - Moved the Python scripts among gnt-* and ganeti-* into modules
459 - Moved all code related to setting up SSH to an external script,
461 - Infrastructure changes for node group support in future versions
467 *(Released Fri, 19 Nov 2010)*
469 A few small bugs fixed, and some improvements to the build system:
471 - Fix documentation regarding conversion to drbd
472 - Fix validation of parameters in cluster modify (``gnt-cluster modify
474 - Fix error handling in node modify with multiple changes
475 - Allow remote imports without checked names
481 *(Released Tue, 19 Oct 2010)*
483 - Disable SSL session ID cache in RPC client
489 *(Released Thu, 14 Oct 2010)*
491 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
492 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
494 - Fix problems with interaction between Curl and Python's HTTP server,
495 resulting in increased speed in many RPC calls
496 - Improve our release script to prevent breakage with older aclocal and
503 *(Released Thu, 7 Oct 2010)*
505 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
506 value retrieved from master
507 - Added support for blacklisted or hidden OS definitions
508 - Added simple lock monitor (accessible via (``gnt-debug locks``)
509 - Added support for -mem-path in KVM hypervisor abstraction layer
510 - Allow overriding instance parameters in tool for inter-cluster
511 instance moves (``tools/move-instance``)
512 - Improved opcode summaries (e.g. in ``gnt-job list``)
513 - Improve consistency of OS listing by sorting it
514 - Documentation updates
520 *(Released Fri, 8 Oct 2010)*
522 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
528 *(Released Mon, 4 Oct 2010)*
530 - Fixed regression in ``gnt-instance rename``
536 *(Released Wed, 22 Sep 2010)*
538 - Fixed OS_VARIANT variable for OS scripts
539 - Fixed cluster tag operations via RAPI
540 - Made ``setup-ssh`` exit with non-zero code if an error occurred
541 - Disabled RAPI CA checks in watcher
547 *(Released Mon, 23 Aug 2010)*
549 - Support DRBD versions of the format "a.b.c.d"
551 - Re-introduce support for usage from multiple threads in RAPI client
552 - Instance renames and modify via RAPI
553 - Work around race condition between processing and archival in job
555 - Mark opcodes following failed one as failed, too
556 - Job field ``lock_status`` was removed due to difficulties making it
557 work with the changed job queue in Ganeti 2.2; a better way to monitor
558 locks is expected for a later 2.2.x release
559 - Fixed dry-run behaviour with many commands
560 - Support ``ssh-agent`` again when adding nodes
561 - Many additional bugfixes
567 *(Released Fri, 30 Jul 2010)*
569 Important change: the internal RPC mechanism between Ganeti nodes has
570 changed from using a home-grown http library (based on the Python base
571 libraries) to use the PycURL library. This requires that PycURL is
572 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
573 against GnuTLS by default. cURL's support for GnuTLS had known issues
574 before cURL 7.21.0 and we recommend using the latest cURL release or
575 linking against OpenSSL. Most other distributions already link PycURL
576 and cURL against OpenSSL. The command::
578 python -c 'import pycurl; print pycurl.version'
580 can be used to determine the libraries PycURL and cURL are linked
583 Other significant changes:
585 - Rewrote much of the internals of the job queue, in order to achieve
586 better parallelism; this decouples job query operations from the job
587 processing, and it should allow much nicer behaviour of the master
588 daemon under load, and it also has uncovered some long-standing bugs
589 related to the job serialisation (now fixed)
590 - Added a default iallocator setting to the cluster parameters,
591 eliminating the need to always pass nodes or an iallocator for
592 operations that require selection of new node(s)
593 - Added experimental support for the LXC virtualization method
594 - Added support for OS parameters, which allows the installation of
595 instances to pass parameter to OS scripts in order to customise the
597 - Added a hypervisor parameter controlling the migration type (live or
598 non-live), since hypervisors have various levels of reliability; this
599 has renamed the 'live' parameter to 'mode'
600 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
601 logical volumes, meaning that cluster verify will ignore them and not
602 flag their presence as errors
603 - The watcher will now reset the error count for failed instances after
604 8 hours, thus allowing self-healing if the problem that caused the
605 instances to be down/fail to start has cleared in the meantime
606 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
607 check for, and warn, if the drbd module parameter ``usermode_helper``
608 is not consistent with the cluster-wide setting; this is needed to
609 make diagnose easier of failed drbd creations
610 - Started adding base IPv6 support, but this is not yet
611 enabled/available for use
612 - Rename operations (cluster, instance) will now return the new name,
613 which is especially useful if a short name was passed in
614 - Added support for instance migration in RAPI
615 - Added a tool to pre-configure nodes for the SSH setup, before joining
616 them to the cluster; this will allow in the future a simplified model
617 for node joining (but not yet fully enabled in 2.2); this needs the
618 paramiko python library
619 - Fixed handling of name-resolving errors
620 - Fixed consistency of job results on the error path
621 - Fixed master-failover race condition when executed multiple times in
623 - Fixed many bugs related to the job queue (mostly introduced during the
624 2.2 development cycle, so not all are impacting 2.1)
625 - Fixed instance migration with missing disk symlinks
626 - Fixed handling of unknown jobs in ``gnt-job archive``
627 - And many other small fixes/improvements
631 - Enhanced both the unittest and the QA coverage
632 - Switched the opcode validation to a generic model, and extended the
633 validation to all opcode parameters
634 - Changed more parts of the code that write shell scripts to use the
636 - Switched the master daemon to use the asyncore library for the Luxi
643 *(Released Thu, 17 Jun 2010)*
645 - Added tool (``move-instance``) and infrastructure to move instances
646 between separate clusters (see :doc:`separate documentation
647 <move-instance>` and :doc:`design document <design-2.2>`)
648 - Added per-request RPC timeout
649 - RAPI now requires a Content-Type header for requests with a body (e.g.
650 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
651 :rfc:`2616` (HTTP/1.1), section 7.2.1)
652 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
654 - Implemented initial support for running Ganeti daemons as separate
655 users, see configure-time flags ``--with-user-prefix`` and
656 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
658 - Instances can be removed after export (``gnt-backup export
660 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
661 key (instead of 1024 bit)
662 - Added new cluster configuration file for cluster domain secret
663 - Import/export now use SSL instead of SSH
664 - Added support for showing estimated time when exporting an instance,
665 see the ``ganeti-os-interface(7)`` manpage and look for
672 *(Released Tue, 16 Nov 2010)*
674 Some more bugfixes. Unless critical bugs occur, this will be the last
677 - Fix case of MAC special-values
678 - Fix mac checker regex
679 - backend: Fix typo causing "out of range" error
680 - Add missing --units in gnt-instance list man page
686 *(Released Tue, 24 Aug 2010)*
689 - Don't ignore secondary node silently on non-mirrored disk templates
691 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
692 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
693 - Properly document the usb_mouse parameter
694 - Fix path in ganeti-rapi(8) (issue 116)
695 - Adjust error message when the ganeti user's .ssh directory is
697 - Add same-node-check when changing the disk template to drbd
703 *(Released Fri, 16 Jul 2010)*
706 - Add an option to only select some reboot types during qa/burnin.
707 (on some hypervisors consequent reboots are not supported)
708 - Fix infrequent race condition in master failover. Sometimes the old
709 master ip address would be still detected as up for a short time
710 after it was removed, causing failover to fail.
711 - Decrease mlockall warnings when the ctypes module is missing. On
712 Python 2.4 we support running even if no ctypes module is installed,
713 but we were too verbose about this issue.
714 - Fix building on old distributions, on which man doesn't have a
716 - Fix RAPI not to ignore the MAC address on instance creation
717 - Implement the old instance creation format in the RAPI client.
723 *(Released Thu, 01 Jul 2010)*
725 A small bugfix release:
726 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
727 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
728 a non-existing option
729 - Fix instance create: the --force-variant option was ignored
730 - Improve pylint 0.21 compatibility and warnings with Python 2.6
731 - Fix modify node storage with non-FQDN arguments
732 - Fix RAPI client to authenticate under Python 2.6 when used
733 for more than 5 requests needing authentication
734 - Fix gnt-instance modify -t (storage) giving a wrong error message
735 when converting a non-shutdown drbd instance to plain
741 *(Released Fri, 18 Jun 2010)*
743 A small bugfix release:
745 - Fix live migration of KVM instances started with older Ganeti
746 versions which had fewer hypervisor parameters
747 - Fix gnt-instance grow-disk on down instances
748 - Fix an error-reporting bug during instance migration
749 - Better checking of the ``--net`` and ``--disk`` values, to avoid
750 silently ignoring broken ones
751 - Fix an RPC error reporting bug affecting, for example, RAPI client
753 - Fix bug triggered by different API version os-es on different nodes
754 - Fix a bug in instance startup with custom hvparams: OS level
755 parameters would fail to be applied.
756 - Fix the RAPI client under Python 2.6 (but more work is needed to
757 make it work completely well with OpenSSL)
758 - Fix handling of errors when resolving names from DNS
764 *(Released Thu, 3 Jun 2010)*
766 A medium sized development cycle. Some new features, and some
767 fixes/small improvements/cleanups.
772 The node deamon now tries to mlock itself into memory, unless the
773 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
774 its logs, and falls back to console logging. This allows emergency
775 features such as ``gnt-node powercycle`` to work even in the event of a
776 broken node disk (tested offlining the disk hosting the node's
777 filesystem and dropping its memory caches; don't try this at home)
779 KVM: add vhost-net acceleration support. It can be tested with a new
780 enough version of the kernel and of qemu-kvm.
782 KVM: Add instance chrooting feature. If you use privilege dropping for
783 your VMs you can also now force them to chroot to an empty directory,
784 before starting the emulated guest.
786 KVM: Add maximum migration bandwith and maximum downtime tweaking
787 support (requires a new-enough version of qemu-kvm).
789 Cluster verify will now warn if the master node doesn't have the master
792 Add a new (incompatible) instance creation request format to RAPI which
793 supports all parameters (previously only a subset was supported, and it
794 wasn't possible to extend the old format to accomodate all the new
795 features. The old format is still supported, and a client can check for
796 this feature, before using it, by checking for its presence in the
797 ``features`` RAPI resource.
799 Now with ancient latin support. Try it passing the ``--roman`` option to
800 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
801 (requires the python-roman module to be installed, in order to work).
806 As usual many internal code refactorings, documentation updates, and
809 - Lots of improvements and cleanups to the experimental Remote API
810 (RAPI) client library.
811 - A new unit test suite for the core daemon libraries.
812 - A fix to creating missing directories makes sure the umask is not
813 applied anymore. This enforces the same directory permissions
815 - Better handling terminating daemons with ctrl+c (used when running
816 them in debugging mode).
817 - Fix a race condition in live migrating a KVM instance, when stat()
818 on the old proc status file returned EINVAL, which is an unexpected
820 - Fixed manpage checking with newer man and utf-8 charachters. But now
821 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
827 *(Released Fri, 7 May 2010)*
829 Fix a bug which prevented untagged KVM instances from starting.
835 *(Released Fri, 7 May 2010)*
837 Another release with a long development cycle, during which many
838 different features were added.
843 The KVM hypervisor now can run the individual instances as non-root, to
844 reduce the impact of a VM being hijacked due to bugs in the
845 hypervisor. It is possible to run all instances as a single (non-root)
846 user, to manually specify a user for each instance, or to dynamically
847 allocate a user out of a cluster-wide pool to each instance, with the
848 guarantee that no two instances will run under the same user ID on any
851 An experimental RAPI client library, that can be used standalone
852 (without the other Ganeti libraries), is provided in the source tree as
853 ``lib/rapi/client.py``. Note this client might change its interface in
854 the future, as we iterate on its capabilities.
856 A new command, ``gnt-cluster renew-crypto`` has been added to easily
857 replace the cluster's certificates and crypto keys. This might help in
858 case they have been compromised, or have simply expired.
860 A new disk option for instance creation has been added that allows one
861 to "adopt" currently existing logical volumes, with data
862 preservation. This should allow easier migration to Ganeti from
863 unmanaged (or managed via other software) instances.
865 Another disk improvement is the possibility to convert between redundant
866 (DRBD) and plain (LVM) disk configuration for an instance. This should
867 allow better scalability (starting with one node and growing the
868 cluster, or shrinking a two-node cluster to one node).
870 A new feature that could help with automated node failovers has been
871 implemented: if a node sees itself as offline (by querying the master
872 candidates), it will try to shutdown (hard) all instances and any active
873 DRBD devices. This reduces the risk of duplicate instances if an
874 external script automatically failovers the instances on such nodes. To
875 enable this, the cluster parameter ``maintain_node_health`` should be
876 enabled; in the future this option (per the name) will enable other
877 automatic maintenance features.
879 Instance export/import now will reuse the original instance
880 specifications for all parameters; that means exporting an instance,
881 deleting it and the importing it back should give an almost identical
882 instance. Note that the default import behaviour has changed from
883 before, where it created only one NIC; now it recreates the original
886 Cluster verify has added a few new checks: SSL certificates validity,
887 /etc/hosts consistency across the cluster, etc.
892 As usual, many internal changes were done, documentation fixes,
895 - Fixed cluster initialization with disabled cluster storage (regression
897 - File-based storage supports growing the disks
898 - Fixed behaviour of node role changes
899 - Fixed cluster verify for some corner cases, plus a general rewrite of
900 cluster verify to allow future extension with more checks
901 - Fixed log spamming by watcher and node daemon (regression introduced
903 - Fixed possible validation issues when changing the list of enabled
905 - Fixed cleanup of /etc/hosts during node removal
906 - Fixed RAPI response for invalid methods
907 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
908 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
909 from ide disks, etc.)
910 - Allow OS changes without re-installation (to record a changed OS
911 outside of Ganeti, or to allow OS renames)
912 - Allow instance creation without OS installation (useful for example if
913 the OS will be installed manually, or restored from a backup not in
915 - Implemented option to make cluster ``copyfile`` use the replication
917 - Added list of enabled hypervisors to ssconf (possibly useful for
919 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
921 - A partial form of node re-IP is possible via node readd, which now
922 allows changed node primary IP
923 - Command line utilities now show an informational message if the job is
925 - The logs of the master daemon now show the PID/UID/GID of the
932 *(Released Fri, 12 Mar 2010)*
934 During the 2.1.0 long release candidate cycle, a lot of improvements and
935 changes have accumulated with were released later as 2.1.1.
940 The node evacuate command (``gnt-node evacuate``) was significantly
941 rewritten, and as such the IAllocator protocol was changed - a new
942 request type has been added. This unfortunate change during a stable
943 series is designed to improve performance of node evacuations; on
944 clusters with more than about five nodes and which are well-balanced,
945 evacuation should proceed in parallel for all instances of the node
946 being evacuated. As such, any existing IAllocator scripts need to be
947 updated, otherwise the above command will fail due to the unknown
948 request. The provided "dumb" allocator has not been updated; but the
949 ganeti-htools package supports the new protocol since version 0.2.4.
951 Another important change is increased validation of node and instance
952 names. This might create problems in special cases, if invalid host
953 names are being used.
955 Also, a new layer of hypervisor parameters has been added, that sits at
956 OS level between the cluster defaults and the instance ones. This allows
957 customisation of virtualization parameters depending on the installed
958 OS. For example instances with OS 'X' may have a different KVM kernel
959 (or any other parameter) than the cluster defaults. This is intended to
960 help managing a multiple OSes on the same cluster, without manual
961 modification of each instance's parameters.
963 A tool for merging clusters, ``cluster-merge``, has been added in the
969 - Improved the int/float conversions that should make the code more
970 robust in face of errors from the node daemons
971 - Fixed the remove node code in case of internal configuration errors
972 - Fixed the node daemon behaviour in face of inconsistent queue
973 directory (e.g. read-only file-system where we can't open the files
975 - Fixed the behaviour of gnt-node modify for master candidate demotion;
976 now it either aborts cleanly or, if given the new "auto_promote"
977 parameter, will automatically promote other nodes as needed
978 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
979 completely prevent Ganeti from working
980 - Fixed bug for instance export when not all disks were successfully
982 - Fixed behaviour of node add when the new node is slow in starting up
984 - Fixed handling of signals in the LUXI client, which should improve
985 behaviour of command-line scripts
986 - Added checks for invalid node/instance names in the configuration (now
987 flagged during cluster verify)
988 - Fixed watcher behaviour for disk activation errors
989 - Fixed two potentially endless loops in http library, which led to the
990 RAPI daemon hanging and consuming 100% CPU in some cases
991 - Fixed bug in RAPI daemon related to hashed passwords
992 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
994 - Enhanced compatibility with non-Debian OSes, but not using absolute
995 path in some commands and allowing customisation of the ssh
996 configuration directory
997 - Fixed possible future issue with new Python versions by abiding to the
998 proper use of ``__slots__`` attribute on classes
999 - Added checks that should prevent directory traversal attacks
1000 - Many documentation fixes based on feedback from users
1005 - Added an "early_release" more for instance replace disks and node
1006 evacuate, where we release locks earlier and thus allow higher
1007 parallelism within the cluster
1008 - Added watcher hooks, intended to allow the watcher to restart other
1009 daemons (e.g. from the ganeti-nbma project), but they can be used of
1010 course for any other purpose
1011 - Added a compile-time disable for DRBD barriers, to increase
1012 performance if the administrator trusts the power supply or the
1013 storage system to not lose writes
1014 - Added the option of using syslog for logging instead of, or in
1015 addition to, Ganeti's own log files
1016 - Removed boot restriction for paravirtual NICs for KVM, recent versions
1017 can indeed boot from a paravirtual NIC
1018 - Added a generic debug level for many operations; while this is not
1019 used widely yet, it allows one to pass the debug value all the way to
1021 - Enhanced the hooks environment for instance moves (failovers,
1022 migrations) where the primary/secondary nodes changed during the
1023 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1024 - Enhanced data validations for many user-supplied values; one important
1025 item is the restrictions imposed on instance and node names, which
1026 might reject some (invalid) host names
1027 - Add a configure-time option to disable file-based storage, if it's not
1028 needed; this allows greater security separation between the master
1029 node and the other nodes from the point of view of the inter-node RPC
1031 - Added user notification in interactive tools if job is waiting in the
1032 job queue or trying to acquire locks
1033 - Added log messages when a job is waiting for locks
1034 - Added filtering by node tags in instance operations which admit
1035 multiple instances (start, stop, reboot, reinstall)
1036 - Added a new tool for cluster mergers, ``cluster-merge``
1037 - Parameters from command line which are of the form ``a=b,c=d`` can now
1038 use backslash escapes to pass in values which contain commas,
1039 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1041 - For KVM, the instance name is the first parameter passed to KVM, so
1042 that it's more visible in the process list
1048 *(Released Tue, 2 Mar 2010)*
1050 Ganeti 2.1 brings many improvements with it. Major changes:
1052 - Added infrastructure to ease automated disk repairs
1053 - Added new daemon to export configuration data in a cheaper way than
1054 using the remote API
1055 - Instance NICs can now be routed instead of being associated with a
1057 - Improved job locking logic to reduce impact of jobs acquiring multiple
1058 locks waiting for other long-running jobs
1060 In-depth implementation details can be found in the Ganeti 2.1 design
1066 - Added chroot hypervisor
1067 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1069 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1070 ``bootloader_path`` and ``bootloader_args``)
1071 - Added the ``use_localtime`` option for the xen-hvm and kvm
1072 hypervisors, and the default value for this has changed to false (in
1073 2.0 xen-hvm always enabled it)
1074 - Added luxi call to submit multiple jobs in one go
1075 - Added cluster initialization option to not modify ``/etc/hosts``
1077 - Added network interface parameters
1078 - Added dry run mode to some LUs
1079 - Added RAPI resources:
1081 - ``/2/instances/[instance_name]/info``
1082 - ``/2/instances/[instance_name]/replace-disks``
1083 - ``/2/nodes/[node_name]/evacuate``
1084 - ``/2/nodes/[node_name]/migrate``
1085 - ``/2/nodes/[node_name]/role``
1086 - ``/2/nodes/[node_name]/storage``
1087 - ``/2/nodes/[node_name]/storage/modify``
1088 - ``/2/nodes/[node_name]/storage/repair``
1090 - Added OpCodes to evacuate or migrate all instances on a node
1091 - Added new command to list storage elements on nodes (``gnt-node
1092 list-storage``) and modify them (``gnt-node modify-storage``)
1093 - Added new ssconf files with master candidate IP address
1094 (``ssconf_master_candidates_ips``), node primary IP address
1095 (``ssconf_node_primary_ips``) and node secondary IP address
1096 (``ssconf_node_secondary_ips``)
1097 - Added ``ganeti-confd`` and a client library to query the Ganeti
1098 configuration via UDP
1099 - Added ability to run hooks after cluster initialization and before
1101 - Added automatic mode for disk replace (``gnt-instance replace-disks
1103 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1104 after catastrophic data-loss
1105 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1107 - Added ``gnt-instance move`` command to move instances
1108 - Added ``gnt-cluster watcher`` command to control watcher
1109 - Added ``gnt-node powercycle`` command to powercycle nodes
1110 - Added new job status field ``lock_status``
1111 - Added parseable error codes to cluster verification (``gnt-cluster
1112 verify --error-codes``) and made output less verbose (use
1113 ``--verbose`` to restore previous behaviour)
1114 - Added UUIDs to the main config entities (cluster, nodes, instances)
1115 - Added support for OS variants
1116 - Added support for hashed passwords in the Ganeti remote API users file
1118 - Added option to specify maximum timeout on instance shutdown
1119 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1120 - Added new helper script to start and stop Ganeti daemons
1121 (``daemon-util``), with the intent to reduce the work necessary to
1122 adjust Ganeti for non-Debian distributions and to start/stop daemons
1124 - Added more unittests
1125 - Fixed critical bug in ganeti-masterd startup
1126 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1127 now customisable at the cluster level for both the KVM and Xen
1128 hypervisors using the new ``migration_port`` parameter
1129 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1130 reinstalling an instance
1131 - Allowed ``@`` in tag names
1132 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1133 - Many documentation updates
1134 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1135 - ``gnt-instance reinstall`` can now reinstall multiple instances
1136 - Updated many command line parameters
1137 - Introduced new OS API version 15
1138 - No longer support a default hypervisor
1139 - Treat virtual LVs as inexistent
1140 - Improved job locking logic to reduce lock contention
1141 - Match instance and node names case insensitively
1142 - Reimplemented bash completion script to be more complete
1149 *(Released Thu, 4 Feb 2010)*
1151 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1152 - Fix a string formatting bug
1153 - Improve safety of the code in some error paths
1154 - Improve data validation in the master of values returned from nodes
1160 *(Released Thu, 17 Dec 2009)*
1162 - Fix security issue due to missing validation of iallocator names; this
1163 allows local and remote execution of arbitrary executables
1164 - Fix failure of gnt-node list during instance removal
1165 - Ship the RAPI documentation in the archive
1171 *(Released Wed, 30 Sep 2009)*
1173 - Fixed many wrong messages
1174 - Fixed a few bugs related to the locking library
1175 - Fixed MAC checking at instance creation time
1176 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1177 - Fixed a few issues related to signal handling in both daemons and
1179 - Fixed the example startup script provided
1180 - Fixed insserv dependencies in the example startup script (patch from
1182 - Fixed handling of drained nodes in the iallocator framework
1183 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1185 - Fixed error related to invalid job IDs in job polling
1186 - Fixed job/opcode persistence on unclean master shutdown
1187 - Fixed handling of partial job processing after unclean master
1189 - Fixed error reporting from LUs, previously all errors were converted
1190 into execution errors
1191 - Fixed error reporting from burnin
1192 - Decreased significantly the memory usage of the job queue
1193 - Optimised slightly multi-job submission
1194 - Optimised slightly opcode loading
1195 - Backported the multi-job submit framework from the development
1196 branch; multi-instance start and stop should be faster
1197 - Added script to clean archived jobs after 21 days; this will reduce
1198 the size of the queue directory
1199 - Added some extra checks in disk size tracking
1200 - Added an example ethers hook script
1201 - Added a cluster parameter that prevents Ganeti from modifying of
1203 - Added more node information to RAPI responses
1204 - Added a ``gnt-job watch`` command that allows following the ouput of a
1206 - Added a bind-address option to ganeti-rapi
1207 - Added more checks to the configuration verify
1208 - Enhanced the burnin script such that some operations can be retried
1210 - Converted instance reinstall to multi-instance model
1216 *(Released Fri, 7 Aug 2009)*
1218 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1219 to allow using the pre-2.0.2 behaviour in activation, if any existing
1220 instances have mismatched disk sizes in the configuration
1221 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1222 any configuration mismatches for disk sizes
1223 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1224 failover to work on two-node clusters
1225 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1227 - Fixed detection of OS script errors in ``gnt-backup export``
1228 - Fixed exit code of ``gnt-backup export``
1234 *(Released Fri, 17 Jul 2009)*
1236 - Added experimental support for stripped logical volumes; this should
1237 enhance performance but comes with a higher complexity in the block
1238 device handling; stripping is only enabled when passing
1239 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1240 affected even in the non-stripped mode
1241 - Improved resiliency against transient failures at the end of DRBD
1242 resyncs, and in general of DRBD resync checks
1243 - Fixed a couple of issues with exports and snapshot errors
1244 - Fixed a couple of issues in instance listing
1245 - Added display of the disk size in ``gnt-instance info``
1246 - Fixed checking for valid OSes in instance creation
1247 - Fixed handling of the "vcpus" parameter in instance listing and in
1248 general of invalid parameters
1249 - Fixed http server library, and thus RAPI, to handle invalid
1250 username/password combinations correctly; this means that now they
1251 report unauthorized for queries too, not only for modifications,
1252 allowing earlier detect of configuration problems
1253 - Added a new "role" node list field, equivalent to the master/master
1254 candidate/drained/offline flags combinations
1255 - Fixed cluster modify and changes of candidate pool size
1256 - Fixed cluster verify error messages for wrong files on regular nodes
1257 - Fixed a couple of issues with node demotion from master candidate role
1258 - Fixed node readd issues
1259 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1260 - Added a new ``--no-voting`` option for masterfailover to fix failover
1261 on two-nodes clusters when the former master node is unreachable
1262 - Added instance reinstall over RAPI
1268 *(Released Tue, 16 Jun 2009)*
1270 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1271 allow re-adding the start in single-user option (regression from 1.2)
1272 - the watcher writes the instance status to a file, to allow monitoring
1273 to report the instance status (from the master) based on cached
1274 results of the watcher's queries; while this can get stale if the
1275 watcher is being locked due to other work on the cluster, this is
1276 still an improvement
1277 - the watcher now also restarts the node daemon and the rapi daemon if
1279 - fixed the watcher to handle full and drained queue cases
1280 - hooks export more instance data in the environment, which helps if
1281 hook scripts need to take action based on the instance's properties
1282 (no longer need to query back into ganeti)
1283 - instance failovers when the instance is stopped do not check for free
1284 RAM, so that failing over a stopped instance is possible in low memory
1286 - rapi uses queries for tags instead of jobs (for less job traffic), and
1287 for cluster tags it won't talk to masterd at all but read them from
1289 - a couple of error handling fixes in RAPI
1290 - drbd handling: improved the error handling of inconsistent disks after
1291 resync to reduce the frequency of "there are some degraded disks for
1292 this instance" messages
1293 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1294 error handling path called a wrong function name)
1300 *(Released Wed, 27 May 2009)*
1302 - no changes from rc5
1305 Version 2.0 release candidate 5
1306 -------------------------------
1308 *(Released Wed, 20 May 2009)*
1310 - fix a couple of bugs (validation, argument checks)
1311 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1312 - some small improvements to RAPI and IAllocator
1313 - make watcher automatically start the master daemon if down
1316 Version 2.0 release candidate 4
1317 -------------------------------
1319 *(Released Mon, 27 Apr 2009)*
1321 - change the OS list to not require locks; this helps with big clusters
1322 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1323 volume group is broken
1324 - ``gnt-instance info``, without any arguments, doesn't run for all
1325 instances anymore; either pass ``--all`` or pass the desired
1326 instances; this helps against mistakes on big clusters where listing
1327 the information for all instances takes a long time
1328 - miscellaneous doc and man pages fixes
1331 Version 2.0 release candidate 3
1332 -------------------------------
1334 *(Released Wed, 8 Apr 2009)*
1336 - Change the internal locking model of some ``gnt-node`` commands, in
1337 order to reduce contention (and blocking of master daemon) when
1338 batching many creation/reinstall jobs
1339 - Fixes to Xen soft reboot
1340 - No longer build documentation at build time, instead distribute it in
1341 the archive, in order to reduce the need for the whole docbook/rst
1345 Version 2.0 release candidate 2
1346 -------------------------------
1348 *(Released Fri, 27 Mar 2009)*
1350 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1351 - Fix watcher startup sequence, improves the behaviour of busy clusters
1352 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1353 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1354 bind address and other places
1355 - Some documentation fixes and updates
1358 Version 2.0 release candidate 1
1359 -------------------------------
1361 *(Released Mon, 2 Mar 2009)*
1363 - More documentation updates, now all docs should be more-or-less
1365 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1367 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1369 - Fix an issue related to $libdir/run/ganeti and cluster creation
1375 *(Released Thu, 19 Feb 2009)*
1377 - Xen PVM and KVM have switched the default value for the instance root
1378 disk to the first partition on the first drive, instead of the whole
1379 drive; this means that the OS installation scripts must be changed
1381 - Man pages have been updated
1382 - RAPI has been switched by default to HTTPS, and the exported functions
1383 should all work correctly
1384 - RAPI v1 has been removed
1385 - Many improvements to the KVM hypervisor
1386 - Block device errors are now better reported
1387 - Many other bugfixes and small improvements
1393 *(Released Mon, 26 Jan 2009)*
1395 - Version 2 is a general rewrite of the code and therefore the
1396 differences are too many to list, see the design document for 2.0 in
1397 the ``doc/`` subdirectory for more details
1398 - In this beta version there is not yet a migration path from 1.2 (there
1399 will be one in the final 2.0 release)
1400 - A few significant changes are:
1402 - all commands are executed by a daemon (``ganeti-masterd``) and the
1403 various ``gnt-*`` commands are just front-ends to it
1404 - all the commands are entered into, and executed from a job queue,
1405 see the ``gnt-job(8)`` manpage
1406 - the RAPI daemon supports read-write operations, secured by basic
1407 HTTP authentication on top of HTTPS
1408 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1409 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1410 to migrate to DRBD 8 first while still running Ganeti 1.2)
1411 - DRBD devices are using statically allocated minor numbers, which
1412 will be assigned to existing instances during the migration process
1413 - there is support for both Xen PVM and Xen HVM instances running on
1415 - KVM virtualization is supported too
1416 - file-based storage has been implemented, which means that it is
1417 possible to run the cluster without LVM and DRBD storage, for
1418 example using a shared filesystem exported from shared storage (and
1419 still have live migration)
1425 *(Released Tue, 13 Jan 2009)*
1427 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1428 - Reuse the old instance mac address by default on instance import, if
1429 the instance name is the same.
1430 - Handle situations in which the node info rpc returns incomplete
1432 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1433 - Improved version of batcher:
1435 - state file support
1436 - instance mac address support
1437 - support for HVM clusters/instances
1439 - Add an option to show the number of cpu sockets and nodes in
1441 - Support OSes that handle more than one version of the OS api (but do
1442 not change the current API in any other way)
1443 - Fix ``gnt-node migrate``
1444 - ``gnt-debug`` man page
1445 - Fixes various more typos and small issues
1446 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1452 *(Released Wed, 24 Sep 2008)*
1454 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1455 type of disk exported to fully virtualized instances.
1456 - provide access to the serial console of HVM instances
1457 - instance auto_balance flag, set by default. If turned off it will
1458 avoid warnings on cluster verify if there is not enough memory to fail
1459 over an instance. in the future it will prevent automatically failing
1460 it over when we will support that.
1461 - batcher tool for instance creation, see ``tools/README.batcher``
1462 - ``gnt-instance reinstall --select-os`` to interactively select a new
1463 operating system when reinstalling an instance.
1464 - when changing the memory amount on instance modify a check has been
1465 added that the instance will be able to start. also warnings are
1466 emitted if the instance will not be able to fail over, if auto_balance
1468 - documentation fixes
1469 - sync fields between ``gnt-instance list/modify/add/import``
1470 - fix a race condition in drbd when the sync speed was set after giving
1471 the device a remote peer.
1477 *(Released Tue, 22 Jul 2008)*
1479 - note: the allowed size and number of tags per object were reduced
1480 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1481 - fixed twisted 8.x compatibility
1482 - fixed ``gnt-instance replace-disks`` with iallocator
1483 - add TCP keepalives on twisted connections to detect restarted nodes
1484 - disk increase support, see ``gnt-instance grow-disk``
1485 - implement bulk node/instance query for RAPI
1486 - add tags in node/instance listing (optional)
1487 - experimental migration (and live migration) support, read the man page
1488 for ``gnt-instance migrate``
1489 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1490 has some small improvements in handling its state file
1496 *(Released Fri, 13 Jun 2008)*
1498 - Experimental readonly, REST-based remote API implementation;
1499 automatically started on master node, TCP port 5080, if enabled by
1500 ``--enable-rapi`` parameter to configure script.
1501 - Instance allocator support. Add and import instance accept a
1502 ``--iallocator`` parameter, and call that instance allocator to decide
1503 which node to use for the instance. The iallocator document describes
1504 what's expected from an allocator script.
1505 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1506 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1507 node is lost there is still enough memory to fail over the instances
1509 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1510 ``gnt-cluster verify``, to check for site-specific compliance. All the
1511 hooks will run, and their output, if any, will be displayed. Any
1512 failing hook will make the verification return an error value.
1513 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1514 primary and secondary interfaces
1515 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1516 that is still declared as part of the cluster and has failed.
1517 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1518 specifying output fields, that just adds the chosen fields to the
1520 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1521 export from the filesystem.
1522 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1523 have been added. Using them you can enable/disable acpi and pae
1524 support, and specify a path for a cd image to be exported to the
1525 instance. These parameters as the name suggest only work on HVM
1527 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1528 PAE support will be set to the previously hardcoded values, but the
1529 (previously hardcoded) path to the CDROM ISO image will be unset and
1530 if required, needs to be set manually with ``gnt-instance modify``
1532 - The address to which an instance's VNC console is bound is now
1533 selectable per-instance, rather than being cluster wide. Of course
1534 this only applies to instances controlled via VNC, so currently just
1535 applies to HVM clusters.
1541 *(Released Mon, 18 Feb 2008)*
1543 - more tweaks to the disk activation code (especially helpful for DRBD)
1544 - change the default ``gnt-instance list`` output format, now there is
1545 one combined status field (see the manpage for the exact values this
1547 - some more fixes for the mac export to hooks change
1548 - make Ganeti not break with DRBD 8.2.x (which changed the version
1549 format in ``/proc/drbd``) (issue 24)
1550 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1551 template, allowing migration from DRBD0.7+MD to DRBD8
1557 *(Released Wed, 30 Jan 2008)*
1559 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1561 - add command aliases infrastructure and a few aliases
1562 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1563 man pages and the ``--help`` option of ``gnt-node
1564 list``/``gnt-instance list``
1565 - fix ``gnt-backup list`` with down nodes (issue 21)
1566 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1567 - fix the dist archive and add a check for including svn/git files in
1569 - some developer-related changes: improve the burnin and the QA suite,
1570 add an upload script for testing during development
1576 *(Released Wed, 16 Jan 2008)*
1578 - experimental HVM support, read the install document, section
1579 "Initializing the cluster"
1580 - allow for the PVM hypervisor per-instance kernel and initrd paths
1581 - add a new command ``gnt-cluster verify-disks`` which uses a new
1582 algorithm to improve the reconnection of the DRBD pairs if the device
1583 on the secondary node has gone away
1584 - make logical volume code auto-activate LVs at disk activation time
1585 - slightly improve the speed of activating disks
1586 - allow specification of the MAC address at instance creation time, and
1587 changing it later via ``gnt-instance modify``
1588 - fix handling of external commands that generate lots of output on
1590 - update documentation with regard to minimum version of DRBD8 supported
1596 *(Released Tue, 4 Dec 2007)*
1598 - Log the ``xm create`` output to the node daemon log on failure (to
1599 help diagnosing the error)
1600 - In debug mode, log all external commands output if failed to the logs
1601 - Change parsing of lvm commands to ignore stderr
1607 *(Released Wed, 28 Nov 2007)*
1609 - Another round of updates to the DRBD 8 code to deal with more failures
1610 in the replace secondary node operation
1611 - Some more logging of failures in disk operations (lvm, drbd)
1612 - A few documentation updates
1619 *(Released Tue, 13 Nov 2007)*
1621 - Change configuration file format from Python's Pickle to JSON.
1622 Upgrading is possible using the cfgupgrade utility.
1623 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1624 faster replace disks and is more stable (DRBD 8 has many improvements
1625 compared to DRBD 0.7)
1626 - Added command line tags support (see man pages for ``gnt-instance``,
1627 ``gnt-node``, ``gnt-cluster``)
1628 - Added instance rename support
1629 - Added multi-instance startup/shutdown
1630 - Added cluster rename support
1631 - Added ``gnt-node evacuate`` to simplify some node operations
1632 - Added instance reboot operation that can speedup reboot as compared to
1634 - Soften the requirement that hostnames are in FQDN format
1635 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1637 - Removed dependency on debian's patched fping that uses the
1638 non-standard ``-S`` option
1639 - Now the OS definitions are searched for in multiple, configurable
1640 paths (easier for distros to package)
1641 - Some changes to the hooks infrastructure (especially the new
1642 post-configuration update hook)
1643 - Other small bugfixes
1645 .. vim: set textwidth=72 syntax=rst :
projects / ganeti-local / blob