9 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
10 resource's ``ip_check`` parameter changed from ``True`` to ``False``
11 to match the underlying LUXI interface
12 - When creating file-based instances via RAPI, the ``file_driver``
13 parameter no longer defaults to ``loop`` and must be specified
19 *(Released Fri, 14 Jan 2011)*
24 - Fixed timezone issues when formatting timestamps
25 - Added support for node groups, available via ``gnt-group`` and other
27 - Added out-of-band framework and management, see :doc:`design
28 document <design-oob>`
29 - Removed support for roman numbers from ``gnt-node list`` and
30 ``gnt-instance list``.
31 - Allowed modification of master network interface via ``gnt-cluster
32 modify --master-netdev``
33 - Accept offline secondaries while shutting down instance disks
34 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
35 - Added support for multiple LVM volume groups
36 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
38 - Added commands to list available fields:
40 - ``gnt-node list-fields``
41 - ``gnt-group list-fields``
42 - ``gnt-instance list-fields``
44 - Updated documentation and man pages
49 - Moved ``rapi_users`` file into separate directory, now named
50 ``…/ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
52 - Added new tool for running commands on many machines,
53 ``tools/ganeti-listrunner``
54 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
55 improving the ``gnt-instance console`` output
56 - Allowed customisation of disk index separator at ``configure`` time
57 - Export node group allocation policy to :doc:`iallocator <iallocator>`
58 - Added support for non-partitioned md disks in ``lvmstrap``
59 - Added script to gracefully power off KVM instances
60 - Split ``utils`` module into smaller parts
61 - Changed query operations to return more detailed information, e.g.
62 whether an information is unavailable due to an offline node. To use
63 this new functionality, the LUXI call ``Query`` must be used. Field
64 information is now stored by the master daemon and can be retrieved
65 using ``QueryFields``. Instances, nodes and groups can also be queried
66 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
67 exposed via RAPI). The following commands make use of this
68 infrastructure change:
71 - ``gnt-group list-fields``
73 - ``gnt-node list-fields``
74 - ``gnt-instance list``
75 - ``gnt-instance list-fields``
81 - New RAPI resources (see :doc:`rapi`):
85 - ``/2/groups/[group_name]``
86 - ``/2/groups/[group_name]/assign-nodes``
87 - ``/2/groups/[group_name]/modify``
88 - ``/2/groups/[group_name]/rename``
89 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
93 - Implemented ``no_install`` for instance creation
94 - Implemented OS parameters for instance reinstallation, allowing
95 use of special settings on reinstallation (e.g. for preserving data)
100 - Added IPv6 support in import/export
101 - Pause DRBD synchronization while wiping disks on instance creation
102 - Updated unittests and QA scripts
103 - Improved network parameters passed to KVM
104 - Converted man pages from docbook to reStructuredText
110 *(Released Mon, 20 Dec 2010)*
112 Released version 2.3.1~rc1 without any changes.
118 *(Released Wed, 1 Dec 2010)*
120 - impexpd: Disable OpenSSL compression in socat if possible (backport
121 from master, commit e90739d625b, see :doc:`installation guide
122 <install-quick>` for details)
123 - Changed unittest coverage report to exclude test scripts
124 - Added script to check version format
130 *(Released Wed, 1 Dec 2010)*
132 Released version 2.3.0~rc1 without any changes.
138 *(Released Fri, 19 Nov 2010)*
140 A number of bugfixes and documentation updates:
142 - Update ganeti-os-interface documentation
143 - Fixed a bug related to duplicate MACs or similar items which should be
145 - Fix breakage in OS state modify
146 - Reinstall instance: disallow offline secondaries (fixes bug related to
147 OS changing but reinstall failing)
148 - plus all the other fixes between 2.2.1 and 2.2.2
154 *(Released Tue, 2 Nov 2010)*
156 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
157 - Fixed master failover race with watcher
158 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
160 - Accept previously stopped instance for export with instance removal
161 - Simplify and extend the environment variables for instance OS scripts
162 - Added new node flags, ``master_capable`` and ``vm_capable``
163 - Added optional instance disk wiping prior during allocation. This is a
164 cluster-wide option and can be set/modified using
165 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
166 - Added IPv6 support, see :doc:`design document <design-2.3>` and
168 - Added a new watcher option (``--ignore-pause``)
169 - Added option to ignore offline node on instance start/stop
170 (``--ignore-offline``)
171 - Allow overriding OS parameters with ``gnt-instance reinstall``
172 - Added ability to change node's secondary IP address using ``gnt-node
174 - Implemented privilege separation for all daemons except
175 ``ganeti-noded``, see ``configure`` options
176 - Complain if an instance's disk is marked faulty in ``gnt-cluster
178 - Implemented job priorities (see ``ganeti(7)`` manpage)
179 - Ignore failures while shutting down instances during failover from
181 - Exit daemon's bootstrap process only once daemon is ready
182 - Export more information via ``LUInstanceQuery``/remote API
183 - Improved documentation, QA and unittests
184 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
185 restart if the file was created or changed
186 - Added LUXI protocol version sent with each request and response,
187 allowing detection of server/client mismatches
188 - Moved the Python scripts among gnt-* and ganeti-* into modules
189 - Moved all code related to setting up SSH to an external script,
191 - Infrastructure changes for node group support in future versions
197 *(Released Fri, 19 Nov 2010)*
199 A few small bugs fixed, and some improvements to the build system:
201 - Fix documentation regarding conversion to drbd
202 - Fix validation of parameters in cluster modify (``gnt-cluster modify
204 - Fix error handling in node modify with multiple changes
205 - Allow remote imports without checked names
211 *(Released Tue, 19 Oct 2010)*
213 - Disable SSL session ID cache in RPC client
219 *(Released Thu, 14 Oct 2010)*
221 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
222 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
224 - Fix problems with interaction between Curl and Python's HTTP server,
225 resulting in increased speed in many RPC calls
226 - Improve our release script to prevent breakage with older aclocal and
233 *(Released Thu, 7 Oct 2010)*
235 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
236 value retrieved from master
237 - Added support for blacklisted or hidden OS definitions
238 - Added simple lock monitor (accessible via (``gnt-debug locks``)
239 - Added support for -mem-path in KVM hypervisor abstraction layer
240 - Allow overriding instance parameters in tool for inter-cluster
241 instance moves (``tools/move-instance``)
242 - Improved opcode summaries (e.g. in ``gnt-job list``)
243 - Improve consistency of OS listing by sorting it
244 - Documentation updates
250 *(Released Fri, 8 Oct 2010)*
252 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
258 *(Released Mon, 4 Oct 2010)*
260 - Fixed regression in ``gnt-instance rename``
266 *(Released Wed, 22 Sep 2010)*
268 - Fixed OS_VARIANT variable for OS scripts
269 - Fixed cluster tag operations via RAPI
270 - Made ``setup-ssh`` exit with non-zero code if an error occurred
271 - Disabled RAPI CA checks in watcher
277 *(Released Mon, 23 Aug 2010)*
279 - Support DRBD versions of the format "a.b.c.d"
281 - Re-introduce support for usage from multiple threads in RAPI client
282 - Instance renames and modify via RAPI
283 - Work around race condition between processing and archival in job
285 - Mark opcodes following failed one as failed, too
286 - Job field ``lock_status`` was removed due to difficulties making it
287 work with the changed job queue in Ganeti 2.2; a better way to monitor
288 locks is expected for a later 2.2.x release
289 - Fixed dry-run behaviour with many commands
290 - Support ``ssh-agent`` again when adding nodes
291 - Many additional bugfixes
297 *(Released Fri, 30 Jul 2010)*
299 Important change: the internal RPC mechanism between Ganeti nodes has
300 changed from using a home-grown http library (based on the Python base
301 libraries) to use the PycURL library. This requires that PycURL is
302 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
303 against GnuTLS by default. cURL's support for GnuTLS had known issues
304 before cURL 7.21.0 and we recommend using the latest cURL release or
305 linking against OpenSSL. Most other distributions already link PycURL
306 and cURL against OpenSSL. The command::
308 python -c 'import pycurl; print pycurl.version'
310 can be used to determine the libraries PycURL and cURL are linked
313 Other significant changes:
315 - Rewrote much of the internals of the job queue, in order to achieve
316 better parallelism; this decouples job query operations from the job
317 processing, and it should allow much nicer behaviour of the master
318 daemon under load, and it also has uncovered some long-standing bugs
319 related to the job serialisation (now fixed)
320 - Added a default iallocator setting to the cluster parameters,
321 eliminating the need to always pass nodes or an iallocator for
322 operations that require selection of new node(s)
323 - Added experimental support for the LXC virtualization method
324 - Added support for OS parameters, which allows the installation of
325 instances to pass parameter to OS scripts in order to customise the
327 - Added a hypervisor parameter controlling the migration type (live or
328 non-live), since hypervisors have various levels of reliability; this
329 has renamed the 'live' parameter to 'mode'
330 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
331 logical volumes, meaning that cluster verify will ignore them and not
332 flag their presence as errors
333 - The watcher will now reset the error count for failed instances after
334 8 hours, thus allowing self-healing if the problem that caused the
335 instances to be down/fail to start has cleared in the meantime
336 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
337 check for, and warn, if the drbd module parameter ``usermode_helper``
338 is not consistent with the cluster-wide setting; this is needed to
339 make diagnose easier of failed drbd creations
340 - Started adding base IPv6 support, but this is not yet
341 enabled/available for use
342 - Rename operations (cluster, instance) will now return the new name,
343 which is especially useful if a short name was passed in
344 - Added support for instance migration in RAPI
345 - Added a tool to pre-configure nodes for the SSH setup, before joining
346 them to the cluster; this will allow in the future a simplified model
347 for node joining (but not yet fully enabled in 2.2); this needs the
348 paramiko python library
349 - Fixed handling of name-resolving errors
350 - Fixed consistency of job results on the error path
351 - Fixed master-failover race condition when executed multiple times in
353 - Fixed many bugs related to the job queue (mostly introduced during the
354 2.2 development cycle, so not all are impacting 2.1)
355 - Fixed instance migration with missing disk symlinks
356 - Fixed handling of unknown jobs in ``gnt-job archive``
357 - And many other small fixes/improvements
361 - Enhanced both the unittest and the QA coverage
362 - Switched the opcode validation to a generic model, and extended the
363 validation to all opcode parameters
364 - Changed more parts of the code that write shell scripts to use the
366 - Switched the master daemon to use the asyncore library for the Luxi
373 *(Released Thu, 17 Jun 2010)*
375 - Added tool (``move-instance``) and infrastructure to move instances
376 between separate clusters (see :doc:`separate documentation
377 <move-instance>` and :doc:`design document <design-2.2>`)
378 - Added per-request RPC timeout
379 - RAPI now requires a Content-Type header for requests with a body (e.g.
380 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
381 :rfc:`2616` (HTTP/1.1), section 7.2.1)
382 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
384 - Implemented initial support for running Ganeti daemons as separate
385 users, see configure-time flags ``--with-user-prefix`` and
386 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
388 - Instances can be removed after export (``gnt-backup export
390 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
391 key (instead of 1024 bit)
392 - Added new cluster configuration file for cluster domain secret
393 - Import/export now use SSL instead of SSH
394 - Added support for showing estimated time when exporting an instance,
395 see the ``ganeti-os-interface(7)`` manpage and look for
402 *(Released Tue, 16 Nov 2010)*
404 Some more bugfixes. Unless critical bugs occur, this will be the last
407 - Fix case of MAC special-values
408 - Fix mac checker regex
409 - backend: Fix typo causing “out of range” error
410 - Add missing --units in gnt-instance list man page
416 *(Released Tue, 24 Aug 2010)*
419 - Don't ignore secondary node silently on non-mirrored disk templates
421 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
422 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
423 - Properly document the usb_mouse parameter
424 - Fix path in ganeti-rapi(8) (issue 116)
425 - Adjust error message when the ganeti user's .ssh directory is
427 - Add same-node-check when changing the disk template to drbd
433 *(Released Fri, 16 Jul 2010)*
436 - Add an option to only select some reboot types during qa/burnin.
437 (on some hypervisors consequent reboots are not supported)
438 - Fix infrequent race condition in master failover. Sometimes the old
439 master ip address would be still detected as up for a short time
440 after it was removed, causing failover to fail.
441 - Decrease mlockall warnings when the ctypes module is missing. On
442 Python 2.4 we support running even if no ctypes module is installed,
443 but we were too verbose about this issue.
444 - Fix building on old distributions, on which man doesn't have a
446 - Fix RAPI not to ignore the MAC address on instance creation
447 - Implement the old instance creation format in the RAPI client.
453 *(Released Thu, 01 Jul 2010)*
455 A small bugfix release:
456 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
457 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
458 a non-existing option
459 - Fix instance create: the --force-variant option was ignored
460 - Improve pylint 0.21 compatibility and warnings with Python 2.6
461 - Fix modify node storage with non-FQDN arguments
462 - Fix RAPI client to authenticate under Python 2.6 when used
463 for more than 5 requests needing authentication
464 - Fix gnt-instance modify -t (storage) giving a wrong error message
465 when converting a non-shutdown drbd instance to plain
471 *(Released Fri, 18 Jun 2010)*
473 A small bugfix release:
475 - Fix live migration of KVM instances started with older Ganeti
476 versions which had fewer hypervisor parameters
477 - Fix gnt-instance grow-disk on down instances
478 - Fix an error-reporting bug during instance migration
479 - Better checking of the ``--net`` and ``--disk`` values, to avoid
480 silently ignoring broken ones
481 - Fix an RPC error reporting bug affecting, for example, RAPI client
483 - Fix bug triggered by different API version os-es on different nodes
484 - Fix a bug in instance startup with custom hvparams: OS level
485 parameters would fail to be applied.
486 - Fix the RAPI client under Python 2.6 (but more work is needed to
487 make it work completely well with OpenSSL)
488 - Fix handling of errors when resolving names from DNS
494 *(Released Thu, 3 Jun 2010)*
496 A medium sized development cycle. Some new features, and some
497 fixes/small improvements/cleanups.
502 The node deamon now tries to mlock itself into memory, unless the
503 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
504 its logs, and falls back to console logging. This allows emergency
505 features such as ``gnt-node powercycle`` to work even in the event of a
506 broken node disk (tested offlining the disk hosting the node's
507 filesystem and dropping its memory caches; don't try this at home)
509 KVM: add vhost-net acceleration support. It can be tested with a new
510 enough version of the kernel and of qemu-kvm.
512 KVM: Add instance chrooting feature. If you use privilege dropping for
513 your VMs you can also now force them to chroot to an empty directory,
514 before starting the emulated guest.
516 KVM: Add maximum migration bandwith and maximum downtime tweaking
517 support (requires a new-enough version of qemu-kvm).
519 Cluster verify will now warn if the master node doesn't have the master
522 Add a new (incompatible) instance creation request format to RAPI which
523 supports all parameters (previously only a subset was supported, and it
524 wasn't possible to extend the old format to accomodate all the new
525 features. The old format is still supported, and a client can check for
526 this feature, before using it, by checking for its presence in the
527 ``features`` RAPI resource.
529 Now with ancient latin support. Try it passing the ``--roman`` option to
530 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
531 (requires the python-roman module to be installed, in order to work).
536 As usual many internal code refactorings, documentation updates, and
539 - Lots of improvements and cleanups to the experimental Remote API
540 (RAPI) client library.
541 - A new unit test suite for the core daemon libraries.
542 - A fix to creating missing directories makes sure the umask is not
543 applied anymore. This enforces the same directory permissions
545 - Better handling terminating daemons with ctrl+c (used when running
546 them in debugging mode).
547 - Fix a race condition in live migrating a KVM instance, when stat()
548 on the old proc status file returned EINVAL, which is an unexpected
550 - Fixed manpage checking with newer man and utf-8 charachters. But now
551 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
557 *(Released Fri, 7 May 2010)*
559 Fix a bug which prevented untagged KVM instances from starting.
565 *(Released Fri, 7 May 2010)*
567 Another release with a long development cycle, during which many
568 different features were added.
573 The KVM hypervisor now can run the individual instances as non-root, to
574 reduce the impact of a VM being hijacked due to bugs in the
575 hypervisor. It is possible to run all instances as a single (non-root)
576 user, to manually specify a user for each instance, or to dynamically
577 allocate a user out of a cluster-wide pool to each instance, with the
578 guarantee that no two instances will run under the same user ID on any
581 An experimental RAPI client library, that can be used standalone
582 (without the other Ganeti libraries), is provided in the source tree as
583 ``lib/rapi/client.py``. Note this client might change its interface in
584 the future, as we iterate on its capabilities.
586 A new command, ``gnt-cluster renew-crypto`` has been added to easily
587 replace the cluster's certificates and crypto keys. This might help in
588 case they have been compromised, or have simply expired.
590 A new disk option for instance creation has been added that allows one
591 to "adopt" currently existing logical volumes, with data
592 preservation. This should allow easier migration to Ganeti from
593 unmanaged (or managed via other software) instances.
595 Another disk improvement is the possibility to convert between redundant
596 (DRBD) and plain (LVM) disk configuration for an instance. This should
597 allow better scalability (starting with one node and growing the
598 cluster, or shrinking a two-node cluster to one node).
600 A new feature that could help with automated node failovers has been
601 implemented: if a node sees itself as offline (by querying the master
602 candidates), it will try to shutdown (hard) all instances and any active
603 DRBD devices. This reduces the risk of duplicate instances if an
604 external script automatically failovers the instances on such nodes. To
605 enable this, the cluster parameter ``maintain_node_health`` should be
606 enabled; in the future this option (per the name) will enable other
607 automatic maintenance features.
609 Instance export/import now will reuse the original instance
610 specifications for all parameters; that means exporting an instance,
611 deleting it and the importing it back should give an almost identical
612 instance. Note that the default import behaviour has changed from
613 before, where it created only one NIC; now it recreates the original
616 Cluster verify has added a few new checks: SSL certificates validity,
617 /etc/hosts consistency across the cluster, etc.
622 As usual, many internal changes were done, documentation fixes,
625 - Fixed cluster initialization with disabled cluster storage (regression
627 - File-based storage supports growing the disks
628 - Fixed behaviour of node role changes
629 - Fixed cluster verify for some corner cases, plus a general rewrite of
630 cluster verify to allow future extension with more checks
631 - Fixed log spamming by watcher and node daemon (regression introduced
633 - Fixed possible validation issues when changing the list of enabled
635 - Fixed cleanup of /etc/hosts during node removal
636 - Fixed RAPI response for invalid methods
637 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
638 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
639 from ide disks, etc.)
640 - Allow OS changes without re-installation (to record a changed OS
641 outside of Ganeti, or to allow OS renames)
642 - Allow instance creation without OS installation (useful for example if
643 the OS will be installed manually, or restored from a backup not in
645 - Implemented option to make cluster ``copyfile`` use the replication
647 - Added list of enabled hypervisors to ssconf (possibly useful for
649 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
651 - A partial form of node re-IP is possible via node readd, which now
652 allows changed node primary IP
653 - Command line utilities now show an informational message if the job is
655 - The logs of the master daemon now show the PID/UID/GID of the
662 *(Released Fri, 12 Mar 2010)*
664 During the 2.1.0 long release candidate cycle, a lot of improvements and
665 changes have accumulated with were released later as 2.1.1.
670 The node evacuate command (``gnt-node evacuate``) was significantly
671 rewritten, and as such the IAllocator protocol was changed - a new
672 request type has been added. This unfortunate change during a stable
673 series is designed to improve performance of node evacuations; on
674 clusters with more than about five nodes and which are well-balanced,
675 evacuation should proceed in parallel for all instances of the node
676 being evacuated. As such, any existing IAllocator scripts need to be
677 updated, otherwise the above command will fail due to the unknown
678 request. The provided "dumb" allocator has not been updated; but the
679 ganeti-htools package supports the new protocol since version 0.2.4.
681 Another important change is increased validation of node and instance
682 names. This might create problems in special cases, if invalid host
683 names are being used.
685 Also, a new layer of hypervisor parameters has been added, that sits at
686 OS level between the cluster defaults and the instance ones. This allows
687 customisation of virtualization parameters depending on the installed
688 OS. For example instances with OS 'X' may have a different KVM kernel
689 (or any other parameter) than the cluster defaults. This is intended to
690 help managing a multiple OSes on the same cluster, without manual
691 modification of each instance's parameters.
693 A tool for merging clusters, ``cluster-merge``, has been added in the
699 - Improved the int/float conversions that should make the code more
700 robust in face of errors from the node daemons
701 - Fixed the remove node code in case of internal configuration errors
702 - Fixed the node daemon behaviour in face of inconsistent queue
703 directory (e.g. read-only file-system where we can't open the files
705 - Fixed the behaviour of gnt-node modify for master candidate demotion;
706 now it either aborts cleanly or, if given the new “auto_promote”
707 parameter, will automatically promote other nodes as needed
708 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
709 completely prevent Ganeti from working
710 - Fixed bug for instance export when not all disks were successfully
712 - Fixed behaviour of node add when the new node is slow in starting up
714 - Fixed handling of signals in the LUXI client, which should improve
715 behaviour of command-line scripts
716 - Added checks for invalid node/instance names in the configuration (now
717 flagged during cluster verify)
718 - Fixed watcher behaviour for disk activation errors
719 - Fixed two potentially endless loops in http library, which led to the
720 RAPI daemon hanging and consuming 100% CPU in some cases
721 - Fixed bug in RAPI daemon related to hashed passwords
722 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
724 - Enhanced compatibility with non-Debian OSes, but not using absolute
725 path in some commands and allowing customisation of the ssh
726 configuration directory
727 - Fixed possible future issue with new Python versions by abiding to the
728 proper use of ``__slots__`` attribute on classes
729 - Added checks that should prevent directory traversal attacks
730 - Many documentation fixes based on feedback from users
735 - Added an “early_release” more for instance replace disks and node
736 evacuate, where we release locks earlier and thus allow higher
737 parallelism within the cluster
738 - Added watcher hooks, intended to allow the watcher to restart other
739 daemons (e.g. from the ganeti-nbma project), but they can be used of
740 course for any other purpose
741 - Added a compile-time disable for DRBD barriers, to increase
742 performance if the administrator trusts the power supply or the
743 storage system to not lose writes
744 - Added the option of using syslog for logging instead of, or in
745 addition to, Ganeti's own log files
746 - Removed boot restriction for paravirtual NICs for KVM, recent versions
747 can indeed boot from a paravirtual NIC
748 - Added a generic debug level for many operations; while this is not
749 used widely yet, it allows one to pass the debug value all the way to
751 - Enhanced the hooks environment for instance moves (failovers,
752 migrations) where the primary/secondary nodes changed during the
753 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
754 - Enhanced data validations for many user-supplied values; one important
755 item is the restrictions imposed on instance and node names, which
756 might reject some (invalid) host names
757 - Add a configure-time option to disable file-based storage, if it's not
758 needed; this allows greater security separation between the master
759 node and the other nodes from the point of view of the inter-node RPC
761 - Added user notification in interactive tools if job is waiting in the
762 job queue or trying to acquire locks
763 - Added log messages when a job is waiting for locks
764 - Added filtering by node tags in instance operations which admit
765 multiple instances (start, stop, reboot, reinstall)
766 - Added a new tool for cluster mergers, ``cluster-merge``
767 - Parameters from command line which are of the form ``a=b,c=d`` can now
768 use backslash escapes to pass in values which contain commas,
769 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
771 - For KVM, the instance name is the first parameter passed to KVM, so
772 that it's more visible in the process list
778 *(Released Tue, 2 Mar 2010)*
780 Ganeti 2.1 brings many improvements with it. Major changes:
782 - Added infrastructure to ease automated disk repairs
783 - Added new daemon to export configuration data in a cheaper way than
785 - Instance NICs can now be routed instead of being associated with a
787 - Improved job locking logic to reduce impact of jobs acquiring multiple
788 locks waiting for other long-running jobs
790 In-depth implementation details can be found in the Ganeti 2.1 design
796 - Added chroot hypervisor
797 - Added more options to xen-hvm hypervisor (``kernel_path`` and
799 - Added more options to xen-pvm hypervisor (``use_bootloader``,
800 ``bootloader_path`` and ``bootloader_args``)
801 - Added the ``use_localtime`` option for the xen-hvm and kvm
802 hypervisors, and the default value for this has changed to false (in
803 2.0 xen-hvm always enabled it)
804 - Added luxi call to submit multiple jobs in one go
805 - Added cluster initialization option to not modify ``/etc/hosts``
807 - Added network interface parameters
808 - Added dry run mode to some LUs
809 - Added RAPI resources:
811 - ``/2/instances/[instance_name]/info``
812 - ``/2/instances/[instance_name]/replace-disks``
813 - ``/2/nodes/[node_name]/evacuate``
814 - ``/2/nodes/[node_name]/migrate``
815 - ``/2/nodes/[node_name]/role``
816 - ``/2/nodes/[node_name]/storage``
817 - ``/2/nodes/[node_name]/storage/modify``
818 - ``/2/nodes/[node_name]/storage/repair``
820 - Added OpCodes to evacuate or migrate all instances on a node
821 - Added new command to list storage elements on nodes (``gnt-node
822 list-storage``) and modify them (``gnt-node modify-storage``)
823 - Added new ssconf files with master candidate IP address
824 (``ssconf_master_candidates_ips``), node primary IP address
825 (``ssconf_node_primary_ips``) and node secondary IP address
826 (``ssconf_node_secondary_ips``)
827 - Added ``ganeti-confd`` and a client library to query the Ganeti
828 configuration via UDP
829 - Added ability to run hooks after cluster initialization and before
831 - Added automatic mode for disk replace (``gnt-instance replace-disks
833 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
834 after catastrophic data-loss
835 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
837 - Added ``gnt-instance move`` command to move instances
838 - Added ``gnt-cluster watcher`` command to control watcher
839 - Added ``gnt-node powercycle`` command to powercycle nodes
840 - Added new job status field ``lock_status``
841 - Added parseable error codes to cluster verification (``gnt-cluster
842 verify --error-codes``) and made output less verbose (use
843 ``--verbose`` to restore previous behaviour)
844 - Added UUIDs to the main config entities (cluster, nodes, instances)
845 - Added support for OS variants
846 - Added support for hashed passwords in the Ganeti remote API users file
848 - Added option to specify maximum timeout on instance shutdown
849 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
850 - Added new helper script to start and stop Ganeti daemons
851 (``daemon-util``), with the intent to reduce the work necessary to
852 adjust Ganeti for non-Debian distributions and to start/stop daemons
854 - Added more unittests
855 - Fixed critical bug in ganeti-masterd startup
856 - Removed the configure-time ``kvm-migration-port`` parameter, this is
857 now customisable at the cluster level for both the KVM and Xen
858 hypervisors using the new ``migration_port`` parameter
859 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
860 reinstalling an instance
861 - Allowed ``@`` in tag names
862 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
863 - Many documentation updates
864 - Distribute hypervisor files on ``gnt-cluster redist-conf``
865 - ``gnt-instance reinstall`` can now reinstall multiple instances
866 - Updated many command line parameters
867 - Introduced new OS API version 15
868 - No longer support a default hypervisor
869 - Treat virtual LVs as inexistent
870 - Improved job locking logic to reduce lock contention
871 - Match instance and node names case insensitively
872 - Reimplemented bash completion script to be more complete
879 *(Released Thu, 4 Feb 2010)*
881 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
882 - Fix a string formatting bug
883 - Improve safety of the code in some error paths
884 - Improve data validation in the master of values returned from nodes
890 *(Released Thu, 17 Dec 2009)*
892 - Fix security issue due to missing validation of iallocator names; this
893 allows local and remote execution of arbitrary executables
894 - Fix failure of gnt-node list during instance removal
895 - Ship the RAPI documentation in the archive
901 *(Released Wed, 30 Sep 2009)*
903 - Fixed many wrong messages
904 - Fixed a few bugs related to the locking library
905 - Fixed MAC checking at instance creation time
906 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
907 - Fixed a few issues related to signal handling in both daemons and
909 - Fixed the example startup script provided
910 - Fixed insserv dependencies in the example startup script (patch from
912 - Fixed handling of drained nodes in the iallocator framework
913 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
915 - Fixed error related to invalid job IDs in job polling
916 - Fixed job/opcode persistence on unclean master shutdown
917 - Fixed handling of partial job processing after unclean master
919 - Fixed error reporting from LUs, previously all errors were converted
920 into execution errors
921 - Fixed error reporting from burnin
922 - Decreased significantly the memory usage of the job queue
923 - Optimised slightly multi-job submission
924 - Optimised slightly opcode loading
925 - Backported the multi-job submit framework from the development
926 branch; multi-instance start and stop should be faster
927 - Added script to clean archived jobs after 21 days; this will reduce
928 the size of the queue directory
929 - Added some extra checks in disk size tracking
930 - Added an example ethers hook script
931 - Added a cluster parameter that prevents Ganeti from modifying of
933 - Added more node information to RAPI responses
934 - Added a ``gnt-job watch`` command that allows following the ouput of a
936 - Added a bind-address option to ganeti-rapi
937 - Added more checks to the configuration verify
938 - Enhanced the burnin script such that some operations can be retried
940 - Converted instance reinstall to multi-instance model
946 *(Released Fri, 7 Aug 2009)*
948 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
949 to allow using the pre-2.0.2 behaviour in activation, if any existing
950 instances have mismatched disk sizes in the configuration
951 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
952 any configuration mismatches for disk sizes
953 - Added ``gnt-master cluste-failover --no-voting`` to allow master
954 failover to work on two-node clusters
955 - Fixed the ``--net`` option of ``gnt-backup import``, which was
957 - Fixed detection of OS script errors in ``gnt-backup export``
958 - Fixed exit code of ``gnt-backup export``
964 *(Released Fri, 17 Jul 2009)*
966 - Added experimental support for stripped logical volumes; this should
967 enhance performance but comes with a higher complexity in the block
968 device handling; stripping is only enabled when passing
969 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
970 affected even in the non-stripped mode
971 - Improved resiliency against transient failures at the end of DRBD
972 resyncs, and in general of DRBD resync checks
973 - Fixed a couple of issues with exports and snapshot errors
974 - Fixed a couple of issues in instance listing
975 - Added display of the disk size in ``gnt-instance info``
976 - Fixed checking for valid OSes in instance creation
977 - Fixed handling of the "vcpus" parameter in instance listing and in
978 general of invalid parameters
979 - Fixed http server library, and thus RAPI, to handle invalid
980 username/password combinations correctly; this means that now they
981 report unauthorized for queries too, not only for modifications,
982 allowing earlier detect of configuration problems
983 - Added a new "role" node list field, equivalent to the master/master
984 candidate/drained/offline flags combinations
985 - Fixed cluster modify and changes of candidate pool size
986 - Fixed cluster verify error messages for wrong files on regular nodes
987 - Fixed a couple of issues with node demotion from master candidate role
988 - Fixed node readd issues
989 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
990 - Added a new ``--no-voting`` option for masterfailover to fix failover
991 on two-nodes clusters when the former master node is unreachable
992 - Added instance reinstall over RAPI
998 *(Released Tue, 16 Jun 2009)*
1000 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1001 allow re-adding the start in single-user option (regression from 1.2)
1002 - the watcher writes the instance status to a file, to allow monitoring
1003 to report the instance status (from the master) based on cached
1004 results of the watcher's queries; while this can get stale if the
1005 watcher is being locked due to other work on the cluster, this is
1006 still an improvement
1007 - the watcher now also restarts the node daemon and the rapi daemon if
1009 - fixed the watcher to handle full and drained queue cases
1010 - hooks export more instance data in the environment, which helps if
1011 hook scripts need to take action based on the instance's properties
1012 (no longer need to query back into ganeti)
1013 - instance failovers when the instance is stopped do not check for free
1014 RAM, so that failing over a stopped instance is possible in low memory
1016 - rapi uses queries for tags instead of jobs (for less job traffic), and
1017 for cluster tags it won't talk to masterd at all but read them from
1019 - a couple of error handling fixes in RAPI
1020 - drbd handling: improved the error handling of inconsistent disks after
1021 resync to reduce the frequency of "there are some degraded disks for
1022 this instance" messages
1023 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1024 error handling path called a wrong function name)
1030 *(Released Wed, 27 May 2009)*
1032 - no changes from rc5
1035 Version 2.0 release candidate 5
1036 -------------------------------
1038 *(Released Wed, 20 May 2009)*
1040 - fix a couple of bugs (validation, argument checks)
1041 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1042 - some small improvements to RAPI and IAllocator
1043 - make watcher automatically start the master daemon if down
1046 Version 2.0 release candidate 4
1047 -------------------------------
1049 *(Released Mon, 27 Apr 2009)*
1051 - change the OS list to not require locks; this helps with big clusters
1052 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1053 volume group is broken
1054 - ``gnt-instance info``, without any arguments, doesn't run for all
1055 instances anymore; either pass ``--all`` or pass the desired
1056 instances; this helps against mistakes on big clusters where listing
1057 the information for all instances takes a long time
1058 - miscellaneous doc and man pages fixes
1061 Version 2.0 release candidate 3
1062 -------------------------------
1064 *(Released Wed, 8 Apr 2009)*
1066 - Change the internal locking model of some ``gnt-node`` commands, in
1067 order to reduce contention (and blocking of master daemon) when
1068 batching many creation/reinstall jobs
1069 - Fixes to Xen soft reboot
1070 - No longer build documentation at build time, instead distribute it in
1071 the archive, in order to reduce the need for the whole docbook/rst
1075 Version 2.0 release candidate 2
1076 -------------------------------
1078 *(Released Fri, 27 Mar 2009)*
1080 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1081 - Fix watcher startup sequence, improves the behaviour of busy clusters
1082 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1083 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1084 bind address and other places
1085 - Some documentation fixes and updates
1088 Version 2.0 release candidate 1
1089 -------------------------------
1091 *(Released Mon, 2 Mar 2009)*
1093 - More documentation updates, now all docs should be more-or-less
1095 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1097 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1099 - Fix an issue related to $libdir/run/ganeti and cluster creation
1105 *(Released Thu, 19 Feb 2009)*
1107 - Xen PVM and KVM have switched the default value for the instance root
1108 disk to the first partition on the first drive, instead of the whole
1109 drive; this means that the OS installation scripts must be changed
1111 - Man pages have been updated
1112 - RAPI has been switched by default to HTTPS, and the exported functions
1113 should all work correctly
1114 - RAPI v1 has been removed
1115 - Many improvements to the KVM hypervisor
1116 - Block device errors are now better reported
1117 - Many other bugfixes and small improvements
1123 *(Released Mon, 26 Jan 2009)*
1125 - Version 2 is a general rewrite of the code and therefore the
1126 differences are too many to list, see the design document for 2.0 in
1127 the ``doc/`` subdirectory for more details
1128 - In this beta version there is not yet a migration path from 1.2 (there
1129 will be one in the final 2.0 release)
1130 - A few significant changes are:
1132 - all commands are executed by a daemon (``ganeti-masterd``) and the
1133 various ``gnt-*`` commands are just front-ends to it
1134 - all the commands are entered into, and executed from a job queue,
1135 see the ``gnt-job(8)`` manpage
1136 - the RAPI daemon supports read-write operations, secured by basic
1137 HTTP authentication on top of HTTPS
1138 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1139 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1140 to migrate to DRBD 8 first while still running Ganeti 1.2)
1141 - DRBD devices are using statically allocated minor numbers, which
1142 will be assigned to existing instances during the migration process
1143 - there is support for both Xen PVM and Xen HVM instances running on
1145 - KVM virtualization is supported too
1146 - file-based storage has been implemented, which means that it is
1147 possible to run the cluster without LVM and DRBD storage, for
1148 example using a shared filesystem exported from shared storage (and
1149 still have live migration)
1155 *(Released Tue, 13 Jan 2009)*
1157 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1158 - Reuse the old instance mac address by default on instance import, if
1159 the instance name is the same.
1160 - Handle situations in which the node info rpc returns incomplete
1162 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1163 - Improved version of batcher:
1165 - state file support
1166 - instance mac address support
1167 - support for HVM clusters/instances
1169 - Add an option to show the number of cpu sockets and nodes in
1171 - Support OSes that handle more than one version of the OS api (but do
1172 not change the current API in any other way)
1173 - Fix ``gnt-node migrate``
1174 - ``gnt-debug`` man page
1175 - Fixes various more typos and small issues
1176 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1182 *(Released Wed, 24 Sep 2008)*
1184 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1185 type of disk exported to fully virtualized instances.
1186 - provide access to the serial console of HVM instances
1187 - instance auto_balance flag, set by default. If turned off it will
1188 avoid warnings on cluster verify if there is not enough memory to fail
1189 over an instance. in the future it will prevent automatically failing
1190 it over when we will support that.
1191 - batcher tool for instance creation, see ``tools/README.batcher``
1192 - ``gnt-instance reinstall --select-os`` to interactively select a new
1193 operating system when reinstalling an instance.
1194 - when changing the memory amount on instance modify a check has been
1195 added that the instance will be able to start. also warnings are
1196 emitted if the instance will not be able to fail over, if auto_balance
1198 - documentation fixes
1199 - sync fields between ``gnt-instance list/modify/add/import``
1200 - fix a race condition in drbd when the sync speed was set after giving
1201 the device a remote peer.
1207 *(Released Tue, 22 Jul 2008)*
1209 - note: the allowed size and number of tags per object were reduced
1210 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1211 - fixed twisted 8.x compatibility
1212 - fixed ``gnt-instance replace-disks`` with iallocator
1213 - add TCP keepalives on twisted connections to detect restarted nodes
1214 - disk increase support, see ``gnt-instance grow-disk``
1215 - implement bulk node/instance query for RAPI
1216 - add tags in node/instance listing (optional)
1217 - experimental migration (and live migration) support, read the man page
1218 for ``gnt-instance migrate``
1219 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1220 has some small improvements in handling its state file
1226 *(Released Fri, 13 Jun 2008)*
1228 - Experimental readonly, REST-based remote API implementation;
1229 automatically started on master node, TCP port 5080, if enabled by
1230 ``--enable-rapi`` parameter to configure script.
1231 - Instance allocator support. Add and import instance accept a
1232 ``--iallocator`` parameter, and call that instance allocator to decide
1233 which node to use for the instance. The iallocator document describes
1234 what's expected from an allocator script.
1235 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1236 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1237 node is lost there is still enough memory to fail over the instances
1239 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1240 ``gnt-cluster verify``, to check for site-specific compliance. All the
1241 hooks will run, and their output, if any, will be displayed. Any
1242 failing hook will make the verification return an error value.
1243 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1244 primary and secondary interfaces
1245 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1246 that is still declared as part of the cluster and has failed.
1247 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1248 specifying output fields, that just adds the chosen fields to the
1250 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1251 export from the filesystem.
1252 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1253 have been added. Using them you can enable/disable acpi and pae
1254 support, and specify a path for a cd image to be exported to the
1255 instance. These parameters as the name suggest only work on HVM
1257 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1258 PAE support will be set to the previously hardcoded values, but the
1259 (previously hardcoded) path to the CDROM ISO image will be unset and
1260 if required, needs to be set manually with ``gnt-instance modify``
1262 - The address to which an instance's VNC console is bound is now
1263 selectable per-instance, rather than being cluster wide. Of course
1264 this only applies to instances controlled via VNC, so currently just
1265 applies to HVM clusters.
1271 *(Released Mon, 18 Feb 2008)*
1273 - more tweaks to the disk activation code (especially helpful for DRBD)
1274 - change the default ``gnt-instance list`` output format, now there is
1275 one combined status field (see the manpage for the exact values this
1277 - some more fixes for the mac export to hooks change
1278 - make Ganeti not break with DRBD 8.2.x (which changed the version
1279 format in ``/proc/drbd``) (issue 24)
1280 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1281 template, allowing migration from DRBD0.7+MD to DRBD8
1287 *(Released Wed, 30 Jan 2008)*
1289 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1291 - add command aliases infrastructure and a few aliases
1292 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1293 man pages and the ``--help`` option of ``gnt-node
1294 list``/``gnt-instance list``
1295 - fix ``gnt-backup list`` with down nodes (issue 21)
1296 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1297 - fix the dist archive and add a check for including svn/git files in
1299 - some developer-related changes: improve the burnin and the QA suite,
1300 add an upload script for testing during development
1306 *(Released Wed, 16 Jan 2008)*
1308 - experimental HVM support, read the install document, section
1309 "Initializing the cluster"
1310 - allow for the PVM hypervisor per-instance kernel and initrd paths
1311 - add a new command ``gnt-cluster verify-disks`` which uses a new
1312 algorithm to improve the reconnection of the DRBD pairs if the device
1313 on the secondary node has gone away
1314 - make logical volume code auto-activate LVs at disk activation time
1315 - slightly improve the speed of activating disks
1316 - allow specification of the MAC address at instance creation time, and
1317 changing it later via ``gnt-instance modify``
1318 - fix handling of external commands that generate lots of output on
1320 - update documentation with regard to minimum version of DRBD8 supported
1326 *(Released Tue, 4 Dec 2007)*
1328 - Log the ``xm create`` output to the node daemon log on failure (to
1329 help diagnosing the error)
1330 - In debug mode, log all external commands output if failed to the logs
1331 - Change parsing of lvm commands to ignore stderr
1337 *(Released Wed, 28 Nov 2007)*
1339 - Another round of updates to the DRBD 8 code to deal with more failures
1340 in the replace secondary node operation
1341 - Some more logging of failures in disk operations (lvm, drbd)
1342 - A few documentation updates
1349 *(Released Tue, 13 Nov 2007)*
1351 - Change configuration file format from Python's Pickle to JSON.
1352 Upgrading is possible using the cfgupgrade utility.
1353 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1354 faster replace disks and is more stable (DRBD 8 has many improvements
1355 compared to DRBD 0.7)
1356 - Added command line tags support (see man pages for ``gnt-instance``,
1357 ``gnt-node``, ``gnt-cluster``)
1358 - Added instance rename support
1359 - Added multi-instance startup/shutdown
1360 - Added cluster rename support
1361 - Added ``gnt-node evacuate`` to simplify some node operations
1362 - Added instance reboot operation that can speedup reboot as compared to
1364 - Soften the requirement that hostnames are in FQDN format
1365 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1367 - Removed dependency on debian's patched fping that uses the
1368 non-standard ``-S`` option
1369 - Now the OS definitions are searched for in multiple, configurable
1370 paths (easier for distros to package)
1371 - Some changes to the hooks infrastructure (especially the new
1372 post-configuration update hook)
1373 - Other small bugfixes
1375 .. vim: set textwidth=72 syntax=rst :
projects / ganeti-local / blob