9 - Moved ``rapi_users`` file into separate directory, now named
10 ``…/ganeti/rapi/users``.
11 - Changed query operations to return more detailed information, e.g.
12 whether an information is unavailable due to an offline node. To use
13 this new functionality, the LUXI call ``Query`` must be used. Field
14 information is now stored by the master daemon and can be retrieved
15 using ``QueryFields``. Instances and nodes can also be queried using
16 the new opcodes ``OpQuery`` and ``OpQueryFields``.
17 - Removed support for roman numbers from ``gnt-node list`` and
18 ``gnt-instance list``.
24 *(Released Wed, 1 Dec 2010)*
26 - impexpd: Disable OpenSSL compression in socat if possible (backport
27 from master, commit e90739d625b, see :doc:`installation guide
28 <install-quick>` for details)
29 - Changed unittest coverage report to exclude test scripts
30 - Added script to check version format
36 *(Released Wed, 1 Dec 2010)*
38 Released version 2.3.0~rc1 without any changes.
44 *(Released Fri, 19 Nov 2010)*
46 A number of bugfixes and documentation updates:
48 - Update ganeti-os-interface documentation
49 - Fixed a bug related to duplicate MACs or similar items which should be
51 - Fix breakage in OS state modify
52 - Reinstall instance: disallow offline secondaries (fixes bug related to
53 OS changing but reinstall failing)
54 - plus all the other fixes between 2.2.1 and 2.2.2
60 *(Released Tue, 2 Nov 2010)*
62 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
63 - Fixed master failover race with watcher
64 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
66 - Accept previously stopped instance for export with instance removal
67 - Simplify and extend the environment variables for instance OS scripts
68 - Added new node flags, ``master_capable`` and ``vm_capable``
69 - Added optional instance disk wiping prior during allocation. This is a
70 cluster-wide option and can be set/modified using
71 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
72 - Added IPv6 support, see :doc:`design document <design-2.3>` and
74 - Added a new watcher option (``--ignore-pause``)
75 - Added option to ignore offline node on instance start/stop
76 (``--ignore-offline``)
77 - Allow overriding OS parameters with ``gnt-instance reinstall``
78 - Added ability to change node's secondary IP address using ``gnt-node
80 - Implemented privilege separation for all daemons except
81 ``ganeti-noded``, see ``configure`` options
82 - Complain if an instance's disk is marked faulty in ``gnt-cluster
84 - Implemented job priorities (see ``ganeti(7)`` manpage)
85 - Ignore failures while shutting down instances during failover from
87 - Exit daemon's bootstrap process only once daemon is ready
88 - Export more information via ``LUQueryInstances``/remote API
89 - Improved documentation, QA and unittests
90 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
91 restart if the file was created or changed
92 - Added LUXI protocol version sent with each request and response,
93 allowing detection of server/client mismatches
94 - Moved the Python scripts among gnt-* and ganeti-* into modules
95 - Moved all code related to setting up SSH to an external script,
97 - Infrastructure changes for node group support in future versions
103 *(Released Fri, 19 Nov 2010)*
105 A few small bugs fixed, and some improvements to the build system:
107 - Fix documentation regarding conversion to drbd
108 - Fix validation of parameters in cluster modify (``gnt-cluster modify
110 - Fix error handling in node modify with multiple changes
111 - Allow remote imports without checked names
117 *(Released Tue, 19 Oct 2010)*
119 - Disable SSL session ID cache in RPC client
125 *(Released Thu, 14 Oct 2010)*
127 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
128 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
130 - Fix problems with interaction between Curl and Python's HTTP server,
131 resulting in increased speed in many RPC calls
132 - Improve our release script to prevent breakage with older aclocal and
139 *(Released Thu, 7 Oct 2010)*
141 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
142 value retrieved from master
143 - Added support for blacklisted or hidden OS definitions
144 - Added simple lock monitor (accessible via (``gnt-debug locks``)
145 - Added support for -mem-path in KVM hypervisor abstraction layer
146 - Allow overriding instance parameters in tool for inter-cluster
147 instance moves (``tools/move-instance``)
148 - Improved opcode summaries (e.g. in ``gnt-job list``)
149 - Improve consistency of OS listing by sorting it
150 - Documentation updates
156 *(Released Fri, 8 Oct 2010)*
158 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
164 *(Released Mon, 4 Oct 2010)*
166 - Fixed regression in ``gnt-instance rename``
172 *(Released Wed, 22 Sep 2010)*
174 - Fixed OS_VARIANT variable for OS scripts
175 - Fixed cluster tag operations via RAPI
176 - Made ``setup-ssh`` exit with non-zero code if an error occurred
177 - Disabled RAPI CA checks in watcher
183 *(Released Mon, 23 Aug 2010)*
185 - Support DRBD versions of the format "a.b.c.d"
187 - Re-introduce support for usage from multiple threads in RAPI client
188 - Instance renames and modify via RAPI
189 - Work around race condition between processing and archival in job
191 - Mark opcodes following failed one as failed, too
192 - Job field ``lock_status`` was removed due to difficulties making it
193 work with the changed job queue in Ganeti 2.2; a better way to monitor
194 locks is expected for a later 2.2.x release
195 - Fixed dry-run behaviour with many commands
196 - Support ``ssh-agent`` again when adding nodes
197 - Many additional bugfixes
203 *(Released Fri, 30 Jul 2010)*
205 Important change: the internal RPC mechanism between Ganeti nodes has
206 changed from using a home-grown http library (based on the Python base
207 libraries) to use the PycURL library. This requires that PycURL is
208 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
209 against GnuTLS by default. cURL's support for GnuTLS had known issues
210 before cURL 7.21.0 and we recommend using the latest cURL release or
211 linking against OpenSSL. Most other distributions already link PycURL
212 and cURL against OpenSSL. The command::
214 python -c 'import pycurl; print pycurl.version'
216 can be used to determine the libraries PycURL and cURL are linked
219 Other significant changes:
221 - Rewrote much of the internals of the job queue, in order to achieve
222 better parallelism; this decouples job query operations from the job
223 processing, and it should allow much nicer behaviour of the master
224 daemon under load, and it also has uncovered some long-standing bugs
225 related to the job serialisation (now fixed)
226 - Added a default iallocator setting to the cluster parameters,
227 eliminating the need to always pass nodes or an iallocator for
228 operations that require selection of new node(s)
229 - Added experimental support for the LXC virtualization method
230 - Added support for OS parameters, which allows the installation of
231 instances to pass parameter to OS scripts in order to customise the
233 - Added a hypervisor parameter controlling the migration type (live or
234 non-live), since hypervisors have various levels of reliability; this
235 has renamed the 'live' parameter to 'mode'
236 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
237 logical volumes, meaning that cluster verify will ignore them and not
238 flag their presence as errors
239 - The watcher will now reset the error count for failed instances after
240 8 hours, thus allowing self-healing if the problem that caused the
241 instances to be down/fail to start has cleared in the meantime
242 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
243 check for, and warn, if the drbd module parameter ``usermode_helper``
244 is not consistent with the cluster-wide setting; this is needed to
245 make diagnose easier of failed drbd creations
246 - Started adding base IPv6 support, but this is not yet
247 enabled/available for use
248 - Rename operations (cluster, instance) will now return the new name,
249 which is especially useful if a short name was passed in
250 - Added support for instance migration in RAPI
251 - Added a tool to pre-configure nodes for the SSH setup, before joining
252 them to the cluster; this will allow in the future a simplified model
253 for node joining (but not yet fully enabled in 2.2); this needs the
254 paramiko python library
255 - Fixed handling of name-resolving errors
256 - Fixed consistency of job results on the error path
257 - Fixed master-failover race condition when executed multiple times in
259 - Fixed many bugs related to the job queue (mostly introduced during the
260 2.2 development cycle, so not all are impacting 2.1)
261 - Fixed instance migration with missing disk symlinks
262 - Fixed handling of unknown jobs in ``gnt-job archive``
263 - And many other small fixes/improvements
267 - Enhanced both the unittest and the QA coverage
268 - Switched the opcode validation to a generic model, and extended the
269 validation to all opcode parameters
270 - Changed more parts of the code that write shell scripts to use the
272 - Switched the master daemon to use the asyncore library for the Luxi
279 *(Released Thu, 17 Jun 2010)*
281 - Added tool (``move-instance``) and infrastructure to move instances
282 between separate clusters (see :doc:`separate documentation
283 <move-instance>` and :doc:`design document <design-2.2>`)
284 - Added per-request RPC timeout
285 - RAPI now requires a Content-Type header for requests with a body (e.g.
286 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
287 :rfc:`2616` (HTTP/1.1), section 7.2.1)
288 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
290 - Implemented initial support for running Ganeti daemons as separate
291 users, see configure-time flags ``--with-user-prefix`` and
292 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
294 - Instances can be removed after export (``gnt-backup export
296 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
297 key (instead of 1024 bit)
298 - Added new cluster configuration file for cluster domain secret
299 - Import/export now use SSL instead of SSH
300 - Added support for showing estimated time when exporting an instance,
301 see the ``ganeti-os-interface(7)`` manpage and look for
308 *(Released Tue, 16 Nov 2010)*
310 Some more bugfixes. Unless critical bugs occur, this will be the last
313 - Fix case of MAC special-values
314 - Fix mac checker regex
315 - backend: Fix typo causing “out of range” error
316 - Add missing --units in gnt-instance list man page
322 *(Released Tue, 24 Aug 2010)*
325 - Don't ignore secondary node silently on non-mirrored disk templates
327 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
328 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
329 - Properly document the usb_mouse parameter
330 - Fix path in ganeti-rapi(8) (issue 116)
331 - Adjust error message when the ganeti user's .ssh directory is
333 - Add same-node-check when changing the disk template to drbd
339 *(Released Fri, 16 Jul 2010)*
342 - Add an option to only select some reboot types during qa/burnin.
343 (on some hypervisors consequent reboots are not supported)
344 - Fix infrequent race condition in master failover. Sometimes the old
345 master ip address would be still detected as up for a short time
346 after it was removed, causing failover to fail.
347 - Decrease mlockall warnings when the ctypes module is missing. On
348 Python 2.4 we support running even if no ctypes module is installed,
349 but we were too verbose about this issue.
350 - Fix building on old distributions, on which man doesn't have a
352 - Fix RAPI not to ignore the MAC address on instance creation
353 - Implement the old instance creation format in the RAPI client.
359 *(Released Thu, 01 Jul 2010)*
361 A small bugfix release:
362 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
363 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
364 a non-existing option
365 - Fix instance create: the --force-variant option was ignored
366 - Improve pylint 0.21 compatibility and warnings with Python 2.6
367 - Fix modify node storage with non-FQDN arguments
368 - Fix RAPI client to authenticate under Python 2.6 when used
369 for more than 5 requests needing authentication
370 - Fix gnt-instance modify -t (storage) giving a wrong error message
371 when converting a non-shutdown drbd instance to plain
377 *(Released Fri, 18 Jun 2010)*
379 A small bugfix release:
381 - Fix live migration of KVM instances started with older Ganeti
382 versions which had fewer hypervisor parameters
383 - Fix gnt-instance grow-disk on down instances
384 - Fix an error-reporting bug during instance migration
385 - Better checking of the ``--net`` and ``--disk`` values, to avoid
386 silently ignoring broken ones
387 - Fix an RPC error reporting bug affecting, for example, RAPI client
389 - Fix bug triggered by different API version os-es on different nodes
390 - Fix a bug in instance startup with custom hvparams: OS level
391 parameters would fail to be applied.
392 - Fix the RAPI client under Python 2.6 (but more work is needed to
393 make it work completely well with OpenSSL)
394 - Fix handling of errors when resolving names from DNS
400 *(Released Thu, 3 Jun 2010)*
402 A medium sized development cycle. Some new features, and some
403 fixes/small improvements/cleanups.
408 The node deamon now tries to mlock itself into memory, unless the
409 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
410 its logs, and falls back to console logging. This allows emergency
411 features such as ``gnt-node powercycle`` to work even in the event of a
412 broken node disk (tested offlining the disk hosting the node's
413 filesystem and dropping its memory caches; don't try this at home)
415 KVM: add vhost-net acceleration support. It can be tested with a new
416 enough version of the kernel and of qemu-kvm.
418 KVM: Add instance chrooting feature. If you use privilege dropping for
419 your VMs you can also now force them to chroot to an empty directory,
420 before starting the emulated guest.
422 KVM: Add maximum migration bandwith and maximum downtime tweaking
423 support (requires a new-enough version of qemu-kvm).
425 Cluster verify will now warn if the master node doesn't have the master
428 Add a new (incompatible) instance creation request format to RAPI which
429 supports all parameters (previously only a subset was supported, and it
430 wasn't possible to extend the old format to accomodate all the new
431 features. The old format is still supported, and a client can check for
432 this feature, before using it, by checking for its presence in the
433 ``features`` RAPI resource.
435 Now with ancient latin support. Try it passing the ``--roman`` option to
436 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
437 (requires the python-roman module to be installed, in order to work).
442 As usual many internal code refactorings, documentation updates, and
445 - Lots of improvements and cleanups to the experimental Remote API
446 (RAPI) client library.
447 - A new unit test suite for the core daemon libraries.
448 - A fix to creating missing directories makes sure the umask is not
449 applied anymore. This enforces the same directory permissions
451 - Better handling terminating daemons with ctrl+c (used when running
452 them in debugging mode).
453 - Fix a race condition in live migrating a KVM instance, when stat()
454 on the old proc status file returned EINVAL, which is an unexpected
456 - Fixed manpage checking with newer man and utf-8 charachters. But now
457 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
463 *(Released Fri, 7 May 2010)*
465 Fix a bug which prevented untagged KVM instances from starting.
471 *(Released Fri, 7 May 2010)*
473 Another release with a long development cycle, during which many
474 different features were added.
479 The KVM hypervisor now can run the individual instances as non-root, to
480 reduce the impact of a VM being hijacked due to bugs in the
481 hypervisor. It is possible to run all instances as a single (non-root)
482 user, to manually specify a user for each instance, or to dynamically
483 allocate a user out of a cluster-wide pool to each instance, with the
484 guarantee that no two instances will run under the same user ID on any
487 An experimental RAPI client library, that can be used standalone
488 (without the other Ganeti libraries), is provided in the source tree as
489 ``lib/rapi/client.py``. Note this client might change its interface in
490 the future, as we iterate on its capabilities.
492 A new command, ``gnt-cluster renew-crypto`` has been added to easily
493 replace the cluster's certificates and crypto keys. This might help in
494 case they have been compromised, or have simply expired.
496 A new disk option for instance creation has been added that allows one
497 to "adopt" currently existing logical volumes, with data
498 preservation. This should allow easier migration to Ganeti from
499 unmanaged (or managed via other software) instances.
501 Another disk improvement is the possibility to convert between redundant
502 (DRBD) and plain (LVM) disk configuration for an instance. This should
503 allow better scalability (starting with one node and growing the
504 cluster, or shrinking a two-node cluster to one node).
506 A new feature that could help with automated node failovers has been
507 implemented: if a node sees itself as offline (by querying the master
508 candidates), it will try to shutdown (hard) all instances and any active
509 DRBD devices. This reduces the risk of duplicate instances if an
510 external script automatically failovers the instances on such nodes. To
511 enable this, the cluster parameter ``maintain_node_health`` should be
512 enabled; in the future this option (per the name) will enable other
513 automatic maintenance features.
515 Instance export/import now will reuse the original instance
516 specifications for all parameters; that means exporting an instance,
517 deleting it and the importing it back should give an almost identical
518 instance. Note that the default import behaviour has changed from
519 before, where it created only one NIC; now it recreates the original
522 Cluster verify has added a few new checks: SSL certificates validity,
523 /etc/hosts consistency across the cluster, etc.
528 As usual, many internal changes were done, documentation fixes,
531 - Fixed cluster initialization with disabled cluster storage (regression
533 - File-based storage supports growing the disks
534 - Fixed behaviour of node role changes
535 - Fixed cluster verify for some corner cases, plus a general rewrite of
536 cluster verify to allow future extension with more checks
537 - Fixed log spamming by watcher and node daemon (regression introduced
539 - Fixed possible validation issues when changing the list of enabled
541 - Fixed cleanup of /etc/hosts during node removal
542 - Fixed RAPI response for invalid methods
543 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
544 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
545 from ide disks, etc.)
546 - Allow OS changes without re-installation (to record a changed OS
547 outside of Ganeti, or to allow OS renames)
548 - Allow instance creation without OS installation (useful for example if
549 the OS will be installed manually, or restored from a backup not in
551 - Implemented option to make cluster ``copyfile`` use the replication
553 - Added list of enabled hypervisors to ssconf (possibly useful for
555 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
557 - A partial form of node re-IP is possible via node readd, which now
558 allows changed node primary IP
559 - Command line utilities now show an informational message if the job is
561 - The logs of the master daemon now show the PID/UID/GID of the
568 *(Released Fri, 12 Mar 2010)*
570 During the 2.1.0 long release candidate cycle, a lot of improvements and
571 changes have accumulated with were released later as 2.1.1.
576 The node evacuate command (``gnt-node evacuate``) was significantly
577 rewritten, and as such the IAllocator protocol was changed - a new
578 request type has been added. This unfortunate change during a stable
579 series is designed to improve performance of node evacuations; on
580 clusters with more than about five nodes and which are well-balanced,
581 evacuation should proceed in parallel for all instances of the node
582 being evacuated. As such, any existing IAllocator scripts need to be
583 updated, otherwise the above command will fail due to the unknown
584 request. The provided "dumb" allocator has not been updated; but the
585 ganeti-htools package supports the new protocol since version 0.2.4.
587 Another important change is increased validation of node and instance
588 names. This might create problems in special cases, if invalid host
589 names are being used.
591 Also, a new layer of hypervisor parameters has been added, that sits at
592 OS level between the cluster defaults and the instance ones. This allows
593 customisation of virtualization parameters depending on the installed
594 OS. For example instances with OS 'X' may have a different KVM kernel
595 (or any other parameter) than the cluster defaults. This is intended to
596 help managing a multiple OSes on the same cluster, without manual
597 modification of each instance's parameters.
599 A tool for merging clusters, ``cluster-merge``, has been added in the
605 - Improved the int/float conversions that should make the code more
606 robust in face of errors from the node daemons
607 - Fixed the remove node code in case of internal configuration errors
608 - Fixed the node daemon behaviour in face of inconsistent queue
609 directory (e.g. read-only file-system where we can't open the files
611 - Fixed the behaviour of gnt-node modify for master candidate demotion;
612 now it either aborts cleanly or, if given the new “auto_promote”
613 parameter, will automatically promote other nodes as needed
614 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
615 completely prevent Ganeti from working
616 - Fixed bug for instance export when not all disks were successfully
618 - Fixed behaviour of node add when the new node is slow in starting up
620 - Fixed handling of signals in the LUXI client, which should improve
621 behaviour of command-line scripts
622 - Added checks for invalid node/instance names in the configuration (now
623 flagged during cluster verify)
624 - Fixed watcher behaviour for disk activation errors
625 - Fixed two potentially endless loops in http library, which led to the
626 RAPI daemon hanging and consuming 100% CPU in some cases
627 - Fixed bug in RAPI daemon related to hashed passwords
628 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
630 - Enhanced compatibility with non-Debian OSes, but not using absolute
631 path in some commands and allowing customisation of the ssh
632 configuration directory
633 - Fixed possible future issue with new Python versions by abiding to the
634 proper use of ``__slots__`` attribute on classes
635 - Added checks that should prevent directory traversal attacks
636 - Many documentation fixes based on feedback from users
641 - Added an “early_release” more for instance replace disks and node
642 evacuate, where we release locks earlier and thus allow higher
643 parallelism within the cluster
644 - Added watcher hooks, intended to allow the watcher to restart other
645 daemons (e.g. from the ganeti-nbma project), but they can be used of
646 course for any other purpose
647 - Added a compile-time disable for DRBD barriers, to increase
648 performance if the administrator trusts the power supply or the
649 storage system to not lose writes
650 - Added the option of using syslog for logging instead of, or in
651 addition to, Ganeti's own log files
652 - Removed boot restriction for paravirtual NICs for KVM, recent versions
653 can indeed boot from a paravirtual NIC
654 - Added a generic debug level for many operations; while this is not
655 used widely yet, it allows one to pass the debug value all the way to
657 - Enhanced the hooks environment for instance moves (failovers,
658 migrations) where the primary/secondary nodes changed during the
659 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
660 - Enhanced data validations for many user-supplied values; one important
661 item is the restrictions imposed on instance and node names, which
662 might reject some (invalid) host names
663 - Add a configure-time option to disable file-based storage, if it's not
664 needed; this allows greater security separation between the master
665 node and the other nodes from the point of view of the inter-node RPC
667 - Added user notification in interactive tools if job is waiting in the
668 job queue or trying to acquire locks
669 - Added log messages when a job is waiting for locks
670 - Added filtering by node tags in instance operations which admit
671 multiple instances (start, stop, reboot, reinstall)
672 - Added a new tool for cluster mergers, ``cluster-merge``
673 - Parameters from command line which are of the form ``a=b,c=d`` can now
674 use backslash escapes to pass in values which contain commas,
675 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
677 - For KVM, the instance name is the first parameter passed to KVM, so
678 that it's more visible in the process list
684 *(Released Tue, 2 Mar 2010)*
686 Ganeti 2.1 brings many improvements with it. Major changes:
688 - Added infrastructure to ease automated disk repairs
689 - Added new daemon to export configuration data in a cheaper way than
691 - Instance NICs can now be routed instead of being associated with a
693 - Improved job locking logic to reduce impact of jobs acquiring multiple
694 locks waiting for other long-running jobs
696 In-depth implementation details can be found in the Ganeti 2.1 design
702 - Added chroot hypervisor
703 - Added more options to xen-hvm hypervisor (``kernel_path`` and
705 - Added more options to xen-pvm hypervisor (``use_bootloader``,
706 ``bootloader_path`` and ``bootloader_args``)
707 - Added the ``use_localtime`` option for the xen-hvm and kvm
708 hypervisors, and the default value for this has changed to false (in
709 2.0 xen-hvm always enabled it)
710 - Added luxi call to submit multiple jobs in one go
711 - Added cluster initialization option to not modify ``/etc/hosts``
713 - Added network interface parameters
714 - Added dry run mode to some LUs
715 - Added RAPI resources:
717 - ``/2/instances/[instance_name]/info``
718 - ``/2/instances/[instance_name]/replace-disks``
719 - ``/2/nodes/[node_name]/evacuate``
720 - ``/2/nodes/[node_name]/migrate``
721 - ``/2/nodes/[node_name]/role``
722 - ``/2/nodes/[node_name]/storage``
723 - ``/2/nodes/[node_name]/storage/modify``
724 - ``/2/nodes/[node_name]/storage/repair``
726 - Added OpCodes to evacuate or migrate all instances on a node
727 - Added new command to list storage elements on nodes (``gnt-node
728 list-storage``) and modify them (``gnt-node modify-storage``)
729 - Added new ssconf files with master candidate IP address
730 (``ssconf_master_candidates_ips``), node primary IP address
731 (``ssconf_node_primary_ips``) and node secondary IP address
732 (``ssconf_node_secondary_ips``)
733 - Added ``ganeti-confd`` and a client library to query the Ganeti
734 configuration via UDP
735 - Added ability to run hooks after cluster initialization and before
737 - Added automatic mode for disk replace (``gnt-instance replace-disks
739 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
740 after catastrophic data-loss
741 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
743 - Added ``gnt-instance move`` command to move instances
744 - Added ``gnt-cluster watcher`` command to control watcher
745 - Added ``gnt-node powercycle`` command to powercycle nodes
746 - Added new job status field ``lock_status``
747 - Added parseable error codes to cluster verification (``gnt-cluster
748 verify --error-codes``) and made output less verbose (use
749 ``--verbose`` to restore previous behaviour)
750 - Added UUIDs to the main config entities (cluster, nodes, instances)
751 - Added support for OS variants
752 - Added support for hashed passwords in the Ganeti remote API users file
754 - Added option to specify maximum timeout on instance shutdown
755 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
756 - Added new helper script to start and stop Ganeti daemons
757 (``daemon-util``), with the intent to reduce the work necessary to
758 adjust Ganeti for non-Debian distributions and to start/stop daemons
760 - Added more unittests
761 - Fixed critical bug in ganeti-masterd startup
762 - Removed the configure-time ``kvm-migration-port`` parameter, this is
763 now customisable at the cluster level for both the KVM and Xen
764 hypervisors using the new ``migration_port`` parameter
765 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
766 reinstalling an instance
767 - Allowed ``@`` in tag names
768 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
769 - Many documentation updates
770 - Distribute hypervisor files on ``gnt-cluster redist-conf``
771 - ``gnt-instance reinstall`` can now reinstall multiple instances
772 - Updated many command line parameters
773 - Introduced new OS API version 15
774 - No longer support a default hypervisor
775 - Treat virtual LVs as inexistent
776 - Improved job locking logic to reduce lock contention
777 - Match instance and node names case insensitively
778 - Reimplemented bash completion script to be more complete
785 *(Released Thu, 4 Feb 2010)*
787 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
788 - Fix a string formatting bug
789 - Improve safety of the code in some error paths
790 - Improve data validation in the master of values returned from nodes
796 *(Released Thu, 17 Dec 2009)*
798 - Fix security issue due to missing validation of iallocator names; this
799 allows local and remote execution of arbitrary executables
800 - Fix failure of gnt-node list during instance removal
801 - Ship the RAPI documentation in the archive
807 *(Released Wed, 30 Sep 2009)*
809 - Fixed many wrong messages
810 - Fixed a few bugs related to the locking library
811 - Fixed MAC checking at instance creation time
812 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
813 - Fixed a few issues related to signal handling in both daemons and
815 - Fixed the example startup script provided
816 - Fixed insserv dependencies in the example startup script (patch from
818 - Fixed handling of drained nodes in the iallocator framework
819 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
821 - Fixed error related to invalid job IDs in job polling
822 - Fixed job/opcode persistence on unclean master shutdown
823 - Fixed handling of partial job processing after unclean master
825 - Fixed error reporting from LUs, previously all errors were converted
826 into execution errors
827 - Fixed error reporting from burnin
828 - Decreased significantly the memory usage of the job queue
829 - Optimised slightly multi-job submission
830 - Optimised slightly opcode loading
831 - Backported the multi-job submit framework from the development
832 branch; multi-instance start and stop should be faster
833 - Added script to clean archived jobs after 21 days; this will reduce
834 the size of the queue directory
835 - Added some extra checks in disk size tracking
836 - Added an example ethers hook script
837 - Added a cluster parameter that prevents Ganeti from modifying of
839 - Added more node information to RAPI responses
840 - Added a ``gnt-job watch`` command that allows following the ouput of a
842 - Added a bind-address option to ganeti-rapi
843 - Added more checks to the configuration verify
844 - Enhanced the burnin script such that some operations can be retried
846 - Converted instance reinstall to multi-instance model
852 *(Released Fri, 7 Aug 2009)*
854 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
855 to allow using the pre-2.0.2 behaviour in activation, if any existing
856 instances have mismatched disk sizes in the configuration
857 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
858 any configuration mismatches for disk sizes
859 - Added ``gnt-master cluste-failover --no-voting`` to allow master
860 failover to work on two-node clusters
861 - Fixed the ``--net`` option of ``gnt-backup import``, which was
863 - Fixed detection of OS script errors in ``gnt-backup export``
864 - Fixed exit code of ``gnt-backup export``
870 *(Released Fri, 17 Jul 2009)*
872 - Added experimental support for stripped logical volumes; this should
873 enhance performance but comes with a higher complexity in the block
874 device handling; stripping is only enabled when passing
875 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
876 affected even in the non-stripped mode
877 - Improved resiliency against transient failures at the end of DRBD
878 resyncs, and in general of DRBD resync checks
879 - Fixed a couple of issues with exports and snapshot errors
880 - Fixed a couple of issues in instance listing
881 - Added display of the disk size in ``gnt-instance info``
882 - Fixed checking for valid OSes in instance creation
883 - Fixed handling of the "vcpus" parameter in instance listing and in
884 general of invalid parameters
885 - Fixed http server library, and thus RAPI, to handle invalid
886 username/password combinations correctly; this means that now they
887 report unauthorized for queries too, not only for modifications,
888 allowing earlier detect of configuration problems
889 - Added a new "role" node list field, equivalent to the master/master
890 candidate/drained/offline flags combinations
891 - Fixed cluster modify and changes of candidate pool size
892 - Fixed cluster verify error messages for wrong files on regular nodes
893 - Fixed a couple of issues with node demotion from master candidate role
894 - Fixed node readd issues
895 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
896 - Added a new ``--no-voting`` option for masterfailover to fix failover
897 on two-nodes clusters when the former master node is unreachable
898 - Added instance reinstall over RAPI
904 *(Released Tue, 16 Jun 2009)*
906 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
907 allow re-adding the start in single-user option (regression from 1.2)
908 - the watcher writes the instance status to a file, to allow monitoring
909 to report the instance status (from the master) based on cached
910 results of the watcher's queries; while this can get stale if the
911 watcher is being locked due to other work on the cluster, this is
913 - the watcher now also restarts the node daemon and the rapi daemon if
915 - fixed the watcher to handle full and drained queue cases
916 - hooks export more instance data in the environment, which helps if
917 hook scripts need to take action based on the instance's properties
918 (no longer need to query back into ganeti)
919 - instance failovers when the instance is stopped do not check for free
920 RAM, so that failing over a stopped instance is possible in low memory
922 - rapi uses queries for tags instead of jobs (for less job traffic), and
923 for cluster tags it won't talk to masterd at all but read them from
925 - a couple of error handling fixes in RAPI
926 - drbd handling: improved the error handling of inconsistent disks after
927 resync to reduce the frequency of "there are some degraded disks for
928 this instance" messages
929 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
930 error handling path called a wrong function name)
936 *(Released Wed, 27 May 2009)*
938 - no changes from rc5
941 Version 2.0 release candidate 5
942 -------------------------------
944 *(Released Wed, 20 May 2009)*
946 - fix a couple of bugs (validation, argument checks)
947 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
948 - some small improvements to RAPI and IAllocator
949 - make watcher automatically start the master daemon if down
952 Version 2.0 release candidate 4
953 -------------------------------
955 *(Released Mon, 27 Apr 2009)*
957 - change the OS list to not require locks; this helps with big clusters
958 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
959 volume group is broken
960 - ``gnt-instance info``, without any arguments, doesn't run for all
961 instances anymore; either pass ``--all`` or pass the desired
962 instances; this helps against mistakes on big clusters where listing
963 the information for all instances takes a long time
964 - miscellaneous doc and man pages fixes
967 Version 2.0 release candidate 3
968 -------------------------------
970 *(Released Wed, 8 Apr 2009)*
972 - Change the internal locking model of some ``gnt-node`` commands, in
973 order to reduce contention (and blocking of master daemon) when
974 batching many creation/reinstall jobs
975 - Fixes to Xen soft reboot
976 - No longer build documentation at build time, instead distribute it in
977 the archive, in order to reduce the need for the whole docbook/rst
981 Version 2.0 release candidate 2
982 -------------------------------
984 *(Released Fri, 27 Mar 2009)*
986 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
987 - Fix watcher startup sequence, improves the behaviour of busy clusters
988 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
989 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
990 bind address and other places
991 - Some documentation fixes and updates
994 Version 2.0 release candidate 1
995 -------------------------------
997 *(Released Mon, 2 Mar 2009)*
999 - More documentation updates, now all docs should be more-or-less
1001 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1003 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1005 - Fix an issue related to $libdir/run/ganeti and cluster creation
1011 *(Released Thu, 19 Feb 2009)*
1013 - Xen PVM and KVM have switched the default value for the instance root
1014 disk to the first partition on the first drive, instead of the whole
1015 drive; this means that the OS installation scripts must be changed
1017 - Man pages have been updated
1018 - RAPI has been switched by default to HTTPS, and the exported functions
1019 should all work correctly
1020 - RAPI v1 has been removed
1021 - Many improvements to the KVM hypervisor
1022 - Block device errors are now better reported
1023 - Many other bugfixes and small improvements
1029 *(Released Mon, 26 Jan 2009)*
1031 - Version 2 is a general rewrite of the code and therefore the
1032 differences are too many to list, see the design document for 2.0 in
1033 the ``doc/`` subdirectory for more details
1034 - In this beta version there is not yet a migration path from 1.2 (there
1035 will be one in the final 2.0 release)
1036 - A few significant changes are:
1038 - all commands are executed by a daemon (``ganeti-masterd``) and the
1039 various ``gnt-*`` commands are just front-ends to it
1040 - all the commands are entered into, and executed from a job queue,
1041 see the ``gnt-job(8)`` manpage
1042 - the RAPI daemon supports read-write operations, secured by basic
1043 HTTP authentication on top of HTTPS
1044 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1045 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1046 to migrate to DRBD 8 first while still running Ganeti 1.2)
1047 - DRBD devices are using statically allocated minor numbers, which
1048 will be assigned to existing instances during the migration process
1049 - there is support for both Xen PVM and Xen HVM instances running on
1051 - KVM virtualization is supported too
1052 - file-based storage has been implemented, which means that it is
1053 possible to run the cluster without LVM and DRBD storage, for
1054 example using a shared filesystem exported from shared storage (and
1055 still have live migration)
1061 *(Released Tue, 13 Jan 2009)*
1063 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1064 - Reuse the old instance mac address by default on instance import, if
1065 the instance name is the same.
1066 - Handle situations in which the node info rpc returns incomplete
1068 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1069 - Improved version of batcher:
1071 - state file support
1072 - instance mac address support
1073 - support for HVM clusters/instances
1075 - Add an option to show the number of cpu sockets and nodes in
1077 - Support OSes that handle more than one version of the OS api (but do
1078 not change the current API in any other way)
1079 - Fix ``gnt-node migrate``
1080 - ``gnt-debug`` man page
1081 - Fixes various more typos and small issues
1082 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1088 *(Released Wed, 24 Sep 2008)*
1090 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1091 type of disk exported to fully virtualized instances.
1092 - provide access to the serial console of HVM instances
1093 - instance auto_balance flag, set by default. If turned off it will
1094 avoid warnings on cluster verify if there is not enough memory to fail
1095 over an instance. in the future it will prevent automatically failing
1096 it over when we will support that.
1097 - batcher tool for instance creation, see ``tools/README.batcher``
1098 - ``gnt-instance reinstall --select-os`` to interactively select a new
1099 operating system when reinstalling an instance.
1100 - when changing the memory amount on instance modify a check has been
1101 added that the instance will be able to start. also warnings are
1102 emitted if the instance will not be able to fail over, if auto_balance
1104 - documentation fixes
1105 - sync fields between ``gnt-instance list/modify/add/import``
1106 - fix a race condition in drbd when the sync speed was set after giving
1107 the device a remote peer.
1113 *(Released Tue, 22 Jul 2008)*
1115 - note: the allowed size and number of tags per object were reduced
1116 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1117 - fixed twisted 8.x compatibility
1118 - fixed ``gnt-instance replace-disks`` with iallocator
1119 - add TCP keepalives on twisted connections to detect restarted nodes
1120 - disk increase support, see ``gnt-instance grow-disk``
1121 - implement bulk node/instance query for RAPI
1122 - add tags in node/instance listing (optional)
1123 - experimental migration (and live migration) support, read the man page
1124 for ``gnt-instance migrate``
1125 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1126 has some small improvements in handling its state file
1132 *(Released Fri, 13 Jun 2008)*
1134 - Experimental readonly, REST-based remote API implementation;
1135 automatically started on master node, TCP port 5080, if enabled by
1136 ``--enable-rapi`` parameter to configure script.
1137 - Instance allocator support. Add and import instance accept a
1138 ``--iallocator`` parameter, and call that instance allocator to decide
1139 which node to use for the instance. The iallocator document describes
1140 what's expected from an allocator script.
1141 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1142 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1143 node is lost there is still enough memory to fail over the instances
1145 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1146 ``gnt-cluster verify``, to check for site-specific compliance. All the
1147 hooks will run, and their output, if any, will be displayed. Any
1148 failing hook will make the verification return an error value.
1149 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1150 primary and secondary interfaces
1151 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1152 that is still declared as part of the cluster and has failed.
1153 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1154 specifying output fields, that just adds the chosen fields to the
1156 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1157 export from the filesystem.
1158 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1159 have been added. Using them you can enable/disable acpi and pae
1160 support, and specify a path for a cd image to be exported to the
1161 instance. These parameters as the name suggest only work on HVM
1163 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1164 PAE support will be set to the previously hardcoded values, but the
1165 (previously hardcoded) path to the CDROM ISO image will be unset and
1166 if required, needs to be set manually with ``gnt-instance modify``
1168 - The address to which an instance's VNC console is bound is now
1169 selectable per-instance, rather than being cluster wide. Of course
1170 this only applies to instances controlled via VNC, so currently just
1171 applies to HVM clusters.
1177 *(Released Mon, 18 Feb 2008)*
1179 - more tweaks to the disk activation code (especially helpful for DRBD)
1180 - change the default ``gnt-instance list`` output format, now there is
1181 one combined status field (see the manpage for the exact values this
1183 - some more fixes for the mac export to hooks change
1184 - make Ganeti not break with DRBD 8.2.x (which changed the version
1185 format in ``/proc/drbd``) (issue 24)
1186 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1187 template, allowing migration from DRBD0.7+MD to DRBD8
1193 *(Released Wed, 30 Jan 2008)*
1195 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1197 - add command aliases infrastructure and a few aliases
1198 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1199 man pages and the ``--help`` option of ``gnt-node
1200 list``/``gnt-instance list``
1201 - fix ``gnt-backup list`` with down nodes (issue 21)
1202 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1203 - fix the dist archive and add a check for including svn/git files in
1205 - some developer-related changes: improve the burnin and the QA suite,
1206 add an upload script for testing during development
1212 *(Released Wed, 16 Jan 2008)*
1214 - experimental HVM support, read the install document, section
1215 "Initializing the cluster"
1216 - allow for the PVM hypervisor per-instance kernel and initrd paths
1217 - add a new command ``gnt-cluster verify-disks`` which uses a new
1218 algorithm to improve the reconnection of the DRBD pairs if the device
1219 on the secondary node has gone away
1220 - make logical volume code auto-activate LVs at disk activation time
1221 - slightly improve the speed of activating disks
1222 - allow specification of the MAC address at instance creation time, and
1223 changing it later via ``gnt-instance modify``
1224 - fix handling of external commands that generate lots of output on
1226 - update documentation with regard to minimum version of DRBD8 supported
1232 *(Released Tue, 4 Dec 2007)*
1234 - Log the ``xm create`` output to the node daemon log on failure (to
1235 help diagnosing the error)
1236 - In debug mode, log all external commands output if failed to the logs
1237 - Change parsing of lvm commands to ignore stderr
1243 *(Released Wed, 28 Nov 2007)*
1245 - Another round of updates to the DRBD 8 code to deal with more failures
1246 in the replace secondary node operation
1247 - Some more logging of failures in disk operations (lvm, drbd)
1248 - A few documentation updates
1255 *(Released Tue, 13 Nov 2007)*
1257 - Change configuration file format from Python's Pickle to JSON.
1258 Upgrading is possible using the cfgupgrade utility.
1259 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1260 faster replace disks and is more stable (DRBD 8 has many improvements
1261 compared to DRBD 0.7)
1262 - Added command line tags support (see man pages for ``gnt-instance``,
1263 ``gnt-node``, ``gnt-cluster``)
1264 - Added instance rename support
1265 - Added multi-instance startup/shutdown
1266 - Added cluster rename support
1267 - Added ``gnt-node evacuate`` to simplify some node operations
1268 - Added instance reboot operation that can speedup reboot as compared to
1270 - Soften the requirement that hostnames are in FQDN format
1271 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1273 - Removed dependency on debian's patched fping that uses the
1274 non-standard ``-S`` option
1275 - Now the OS definitions are searched for in multiple, configurable
1276 paths (easier for distros to package)
1277 - Some changes to the hooks infrastructure (especially the new
1278 post-configuration update hook)
1279 - Other small bugfixes
1281 .. vim: set textwidth=72 syntax=rst :
projects / ganeti-local / blob