7 *(Released Wed, 1 Dec 2010)*
9 - impexpd: Disable OpenSSL compression in socat if possible (backport
10 from master, commit e90739d625b, see :doc:`installation guide
11 <install-quick>` for details)
12 - Changed unittest coverage report to exclude test scripts
13 - Added script to check version format
19 *(Released Wed, 1 Dec 2010)*
21 Released version 2.3.0~rc1 without any changes.
27 *(Released Fri, 19 Nov 2010)*
29 A number of bugfixes and documentation updates:
31 - Update ganeti-os-interface documentation
32 - Fixed a bug related to duplicate MACs or similar items which should be
34 - Fix breakage in OS state modify
35 - Reinstall instance: disallow offline secondaries (fixes bug related to
36 OS changing but reinstall failing)
37 - plus all the other fixes between 2.2.1 and 2.2.2
43 *(Released Tue, 2 Nov 2010)*
45 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
46 - Fixed master failover race with watcher
47 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
49 - Accept previously stopped instance for export with instance removal
50 - Simplify and extend the environment variables for instance OS scripts
51 - Added new node flags, ``master_capable`` and ``vm_capable``
52 - Added optional instance disk wiping prior during allocation. This is a
53 cluster-wide option and can be set/modified using
54 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
55 - Added IPv6 support, see :doc:`design document <design-2.3>` and
57 - Added a new watcher option (``--ignore-pause``)
58 - Added option to ignore offline node on instance start/stop
59 (``--ignore-offline``)
60 - Allow overriding OS parameters with ``gnt-instance reinstall``
61 - Added ability to change node's secondary IP address using ``gnt-node
63 - Implemented privilege separation for all daemons except
64 ``ganeti-noded``, see ``configure`` options
65 - Complain if an instance's disk is marked faulty in ``gnt-cluster
67 - Implemented job priorities (see ``ganeti(7)`` manpage)
68 - Ignore failures while shutting down instances during failover from
70 - Exit daemon's bootstrap process only once daemon is ready
71 - Export more information via ``LUQueryInstances``/remote API
72 - Improved documentation, QA and unittests
73 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
74 restart if the file was created or changed
75 - Added LUXI protocol version sent with each request and response,
76 allowing detection of server/client mismatches
77 - Moved the Python scripts among gnt-* and ganeti-* into modules
78 - Moved all code related to setting up SSH to an external script,
80 - Infrastructure changes for node group support in future versions
86 *(Released Fri, 19 Nov 2010)*
88 A few small bugs fixed, and some improvements to the build system:
90 - Fix documentation regarding conversion to drbd
91 - Fix validation of parameters in cluster modify (``gnt-cluster modify
93 - Fix error handling in node modify with multiple changes
94 - Allow remote imports without checked names
100 *(Released Tue, 19 Oct 2010)*
102 - Disable SSL session ID cache in RPC client
108 *(Released Thu, 14 Oct 2010)*
110 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
111 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
113 - Fix problems with interaction between Curl and Python's HTTP server,
114 resulting in increased speed in many RPC calls
115 - Improve our release script to prevent breakage with older aclocal and
122 *(Released Thu, 7 Oct 2010)*
124 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
125 value retrieved from master
126 - Added support for blacklisted or hidden OS definitions
127 - Added simple lock monitor (accessible via (``gnt-debug locks``)
128 - Added support for -mem-path in KVM hypervisor abstraction layer
129 - Allow overriding instance parameters in tool for inter-cluster
130 instance moves (``tools/move-instance``)
131 - Improved opcode summaries (e.g. in ``gnt-job list``)
132 - Improve consistency of OS listing by sorting it
133 - Documentation updates
139 *(Released Fri, 8 Oct 2010)*
141 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
147 *(Released Mon, 4 Oct 2010)*
149 - Fixed regression in ``gnt-instance rename``
155 *(Released Wed, 22 Sep 2010)*
157 - Fixed OS_VARIANT variable for OS scripts
158 - Fixed cluster tag operations via RAPI
159 - Made ``setup-ssh`` exit with non-zero code if an error occurred
160 - Disabled RAPI CA checks in watcher
166 *(Released Mon, 23 Aug 2010)*
168 - Support DRBD versions of the format "a.b.c.d"
170 - Re-introduce support for usage from multiple threads in RAPI client
171 - Instance renames and modify via RAPI
172 - Work around race condition between processing and archival in job
174 - Mark opcodes following failed one as failed, too
175 - Job field ``lock_status`` was removed due to difficulties making it
176 work with the changed job queue in Ganeti 2.2; a better way to monitor
177 locks is expected for a later 2.2.x release
178 - Fixed dry-run behaviour with many commands
179 - Support ``ssh-agent`` again when adding nodes
180 - Many additional bugfixes
186 *(Released Fri, 30 Jul 2010)*
188 Important change: the internal RPC mechanism between Ganeti nodes has
189 changed from using a home-grown http library (based on the Python base
190 libraries) to use the PycURL library. This requires that PycURL is
191 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
192 against GnuTLS by default. cURL's support for GnuTLS had known issues
193 before cURL 7.21.0 and we recommend using the latest cURL release or
194 linking against OpenSSL. Most other distributions already link PycURL
195 and cURL against OpenSSL. The command::
197 python -c 'import pycurl; print pycurl.version'
199 can be used to determine the libraries PycURL and cURL are linked
202 Other significant changes:
204 - Rewrote much of the internals of the job queue, in order to achieve
205 better parallelism; this decouples job query operations from the job
206 processing, and it should allow much nicer behaviour of the master
207 daemon under load, and it also has uncovered some long-standing bugs
208 related to the job serialisation (now fixed)
209 - Added a default iallocator setting to the cluster parameters,
210 eliminating the need to always pass nodes or an iallocator for
211 operations that require selection of new node(s)
212 - Added experimental support for the LXC virtualization method
213 - Added support for OS parameters, which allows the installation of
214 instances to pass parameter to OS scripts in order to customise the
216 - Added a hypervisor parameter controlling the migration type (live or
217 non-live), since hypervisors have various levels of reliability; this
218 has renamed the 'live' parameter to 'mode'
219 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
220 logical volumes, meaning that cluster verify will ignore them and not
221 flag their presence as errors
222 - The watcher will now reset the error count for failed instances after
223 8 hours, thus allowing self-healing if the problem that caused the
224 instances to be down/fail to start has cleared in the meantime
225 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
226 check for, and warn, if the drbd module parameter ``usermode_helper``
227 is not consistent with the cluster-wide setting; this is needed to
228 make diagnose easier of failed drbd creations
229 - Started adding base IPv6 support, but this is not yet
230 enabled/available for use
231 - Rename operations (cluster, instance) will now return the new name,
232 which is especially useful if a short name was passed in
233 - Added support for instance migration in RAPI
234 - Added a tool to pre-configure nodes for the SSH setup, before joining
235 them to the cluster; this will allow in the future a simplified model
236 for node joining (but not yet fully enabled in 2.2); this needs the
237 paramiko python library
238 - Fixed handling of name-resolving errors
239 - Fixed consistency of job results on the error path
240 - Fixed master-failover race condition when executed multiple times in
242 - Fixed many bugs related to the job queue (mostly introduced during the
243 2.2 development cycle, so not all are impacting 2.1)
244 - Fixed instance migration with missing disk symlinks
245 - Fixed handling of unknown jobs in ``gnt-job archive``
246 - And many other small fixes/improvements
250 - Enhanced both the unittest and the QA coverage
251 - Switched the opcode validation to a generic model, and extended the
252 validation to all opcode parameters
253 - Changed more parts of the code that write shell scripts to use the
255 - Switched the master daemon to use the asyncore library for the Luxi
262 *(Released Thu, 17 Jun 2010)*
264 - Added tool (``move-instance``) and infrastructure to move instances
265 between separate clusters (see :doc:`separate documentation
266 <move-instance>` and :doc:`design document <design-2.2>`)
267 - Added per-request RPC timeout
268 - RAPI now requires a Content-Type header for requests with a body (e.g.
269 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
270 :rfc:`2616` (HTTP/1.1), section 7.2.1)
271 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
273 - Implemented initial support for running Ganeti daemons as separate
274 users, see configure-time flags ``--with-user-prefix`` and
275 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
277 - Instances can be removed after export (``gnt-backup export
279 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
280 key (instead of 1024 bit)
281 - Added new cluster configuration file for cluster domain secret
282 - Import/export now use SSL instead of SSH
283 - Added support for showing estimated time when exporting an instance,
284 see the ``ganeti-os-interface(7)`` manpage and look for
291 *(Released Tue, 16 Nov 2010)*
293 Some more bugfixes. Unless critical bugs occur, this will be the last
296 - Fix case of MAC special-values
297 - Fix mac checker regex
298 - backend: Fix typo causing “out of range” error
299 - Add missing --units in gnt-instance list man page
305 *(Released Tue, 24 Aug 2010)*
308 - Don't ignore secondary node silently on non-mirrored disk templates
310 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
311 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
312 - Properly document the usb_mouse parameter
313 - Fix path in ganeti-rapi(8) (issue 116)
314 - Adjust error message when the ganeti user's .ssh directory is
316 - Add same-node-check when changing the disk template to drbd
322 *(Released Fri, 16 Jul 2010)*
325 - Add an option to only select some reboot types during qa/burnin.
326 (on some hypervisors consequent reboots are not supported)
327 - Fix infrequent race condition in master failover. Sometimes the old
328 master ip address would be still detected as up for a short time
329 after it was removed, causing failover to fail.
330 - Decrease mlockall warnings when the ctypes module is missing. On
331 Python 2.4 we support running even if no ctypes module is installed,
332 but we were too verbose about this issue.
333 - Fix building on old distributions, on which man doesn't have a
335 - Fix RAPI not to ignore the MAC address on instance creation
336 - Implement the old instance creation format in the RAPI client.
342 *(Released Thu, 01 Jul 2010)*
344 A small bugfix release:
345 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
346 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
347 a non-existing option
348 - Fix instance create: the --force-variant option was ignored
349 - Improve pylint 0.21 compatibility and warnings with Python 2.6
350 - Fix modify node storage with non-FQDN arguments
351 - Fix RAPI client to authenticate under Python 2.6 when used
352 for more than 5 requests needing authentication
353 - Fix gnt-instance modify -t (storage) giving a wrong error message
354 when converting a non-shutdown drbd instance to plain
360 *(Released Fri, 18 Jun 2010)*
362 A small bugfix release:
364 - Fix live migration of KVM instances started with older Ganeti
365 versions which had fewer hypervisor parameters
366 - Fix gnt-instance grow-disk on down instances
367 - Fix an error-reporting bug during instance migration
368 - Better checking of the ``--net`` and ``--disk`` values, to avoid
369 silently ignoring broken ones
370 - Fix an RPC error reporting bug affecting, for example, RAPI client
372 - Fix bug triggered by different API version os-es on different nodes
373 - Fix a bug in instance startup with custom hvparams: OS level
374 parameters would fail to be applied.
375 - Fix the RAPI client under Python 2.6 (but more work is needed to
376 make it work completely well with OpenSSL)
377 - Fix handling of errors when resolving names from DNS
383 *(Released Thu, 3 Jun 2010)*
385 A medium sized development cycle. Some new features, and some
386 fixes/small improvements/cleanups.
391 The node deamon now tries to mlock itself into memory, unless the
392 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
393 its logs, and falls back to console logging. This allows emergency
394 features such as ``gnt-node powercycle`` to work even in the event of a
395 broken node disk (tested offlining the disk hosting the node's
396 filesystem and dropping its memory caches; don't try this at home)
398 KVM: add vhost-net acceleration support. It can be tested with a new
399 enough version of the kernel and of qemu-kvm.
401 KVM: Add instance chrooting feature. If you use privilege dropping for
402 your VMs you can also now force them to chroot to an empty directory,
403 before starting the emulated guest.
405 KVM: Add maximum migration bandwith and maximum downtime tweaking
406 support (requires a new-enough version of qemu-kvm).
408 Cluster verify will now warn if the master node doesn't have the master
411 Add a new (incompatible) instance creation request format to RAPI which
412 supports all parameters (previously only a subset was supported, and it
413 wasn't possible to extend the old format to accomodate all the new
414 features. The old format is still supported, and a client can check for
415 this feature, before using it, by checking for its presence in the
416 ``features`` RAPI resource.
418 Now with ancient latin support. Try it passing the ``--roman`` option to
419 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
420 (requires the python-roman module to be installed, in order to work).
425 As usual many internal code refactorings, documentation updates, and
428 - Lots of improvements and cleanups to the experimental Remote API
429 (RAPI) client library.
430 - A new unit test suite for the core daemon libraries.
431 - A fix to creating missing directories makes sure the umask is not
432 applied anymore. This enforces the same directory permissions
434 - Better handling terminating daemons with ctrl+c (used when running
435 them in debugging mode).
436 - Fix a race condition in live migrating a KVM instance, when stat()
437 on the old proc status file returned EINVAL, which is an unexpected
439 - Fixed manpage checking with newer man and utf-8 charachters. But now
440 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
446 *(Released Fri, 7 May 2010)*
448 Fix a bug which prevented untagged KVM instances from starting.
454 *(Released Fri, 7 May 2010)*
456 Another release with a long development cycle, during which many
457 different features were added.
462 The KVM hypervisor now can run the individual instances as non-root, to
463 reduce the impact of a VM being hijacked due to bugs in the
464 hypervisor. It is possible to run all instances as a single (non-root)
465 user, to manually specify a user for each instance, or to dynamically
466 allocate a user out of a cluster-wide pool to each instance, with the
467 guarantee that no two instances will run under the same user ID on any
470 An experimental RAPI client library, that can be used standalone
471 (without the other Ganeti libraries), is provided in the source tree as
472 ``lib/rapi/client.py``. Note this client might change its interface in
473 the future, as we iterate on its capabilities.
475 A new command, ``gnt-cluster renew-crypto`` has been added to easily
476 replace the cluster's certificates and crypto keys. This might help in
477 case they have been compromised, or have simply expired.
479 A new disk option for instance creation has been added that allows one
480 to "adopt" currently existing logical volumes, with data
481 preservation. This should allow easier migration to Ganeti from
482 unmanaged (or managed via other software) instances.
484 Another disk improvement is the possibility to convert between redundant
485 (DRBD) and plain (LVM) disk configuration for an instance. This should
486 allow better scalability (starting with one node and growing the
487 cluster, or shrinking a two-node cluster to one node).
489 A new feature that could help with automated node failovers has been
490 implemented: if a node sees itself as offline (by querying the master
491 candidates), it will try to shutdown (hard) all instances and any active
492 DRBD devices. This reduces the risk of duplicate instances if an
493 external script automatically failovers the instances on such nodes. To
494 enable this, the cluster parameter ``maintain_node_health`` should be
495 enabled; in the future this option (per the name) will enable other
496 automatic maintenance features.
498 Instance export/import now will reuse the original instance
499 specifications for all parameters; that means exporting an instance,
500 deleting it and the importing it back should give an almost identical
501 instance. Note that the default import behaviour has changed from
502 before, where it created only one NIC; now it recreates the original
505 Cluster verify has added a few new checks: SSL certificates validity,
506 /etc/hosts consistency across the cluster, etc.
511 As usual, many internal changes were done, documentation fixes,
514 - Fixed cluster initialization with disabled cluster storage (regression
516 - File-based storage supports growing the disks
517 - Fixed behaviour of node role changes
518 - Fixed cluster verify for some corner cases, plus a general rewrite of
519 cluster verify to allow future extension with more checks
520 - Fixed log spamming by watcher and node daemon (regression introduced
522 - Fixed possible validation issues when changing the list of enabled
524 - Fixed cleanup of /etc/hosts during node removal
525 - Fixed RAPI response for invalid methods
526 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
527 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
528 from ide disks, etc.)
529 - Allow OS changes without re-installation (to record a changed OS
530 outside of Ganeti, or to allow OS renames)
531 - Allow instance creation without OS installation (useful for example if
532 the OS will be installed manually, or restored from a backup not in
534 - Implemented option to make cluster ``copyfile`` use the replication
536 - Added list of enabled hypervisors to ssconf (possibly useful for
538 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
540 - A partial form of node re-IP is possible via node readd, which now
541 allows changed node primary IP
542 - Command line utilities now show an informational message if the job is
544 - The logs of the master daemon now show the PID/UID/GID of the
551 *(Released Fri, 12 Mar 2010)*
553 During the 2.1.0 long release candidate cycle, a lot of improvements and
554 changes have accumulated with were released later as 2.1.1.
559 The node evacuate command (``gnt-node evacuate``) was significantly
560 rewritten, and as such the IAllocator protocol was changed - a new
561 request type has been added. This unfortunate change during a stable
562 series is designed to improve performance of node evacuations; on
563 clusters with more than about five nodes and which are well-balanced,
564 evacuation should proceed in parallel for all instances of the node
565 being evacuated. As such, any existing IAllocator scripts need to be
566 updated, otherwise the above command will fail due to the unknown
567 request. The provided "dumb" allocator has not been updated; but the
568 ganeti-htools package supports the new protocol since version 0.2.4.
570 Another important change is increased validation of node and instance
571 names. This might create problems in special cases, if invalid host
572 names are being used.
574 Also, a new layer of hypervisor parameters has been added, that sits at
575 OS level between the cluster defaults and the instance ones. This allows
576 customisation of virtualization parameters depending on the installed
577 OS. For example instances with OS 'X' may have a different KVM kernel
578 (or any other parameter) than the cluster defaults. This is intended to
579 help managing a multiple OSes on the same cluster, without manual
580 modification of each instance's parameters.
582 A tool for merging clusters, ``cluster-merge``, has been added in the
588 - Improved the int/float conversions that should make the code more
589 robust in face of errors from the node daemons
590 - Fixed the remove node code in case of internal configuration errors
591 - Fixed the node daemon behaviour in face of inconsistent queue
592 directory (e.g. read-only file-system where we can't open the files
594 - Fixed the behaviour of gnt-node modify for master candidate demotion;
595 now it either aborts cleanly or, if given the new “auto_promote”
596 parameter, will automatically promote other nodes as needed
597 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
598 completely prevent Ganeti from working
599 - Fixed bug for instance export when not all disks were successfully
601 - Fixed behaviour of node add when the new node is slow in starting up
603 - Fixed handling of signals in the LUXI client, which should improve
604 behaviour of command-line scripts
605 - Added checks for invalid node/instance names in the configuration (now
606 flagged during cluster verify)
607 - Fixed watcher behaviour for disk activation errors
608 - Fixed two potentially endless loops in http library, which led to the
609 RAPI daemon hanging and consuming 100% CPU in some cases
610 - Fixed bug in RAPI daemon related to hashed passwords
611 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
613 - Enhanced compatibility with non-Debian OSes, but not using absolute
614 path in some commands and allowing customisation of the ssh
615 configuration directory
616 - Fixed possible future issue with new Python versions by abiding to the
617 proper use of ``__slots__`` attribute on classes
618 - Added checks that should prevent directory traversal attacks
619 - Many documentation fixes based on feedback from users
624 - Added an “early_release” more for instance replace disks and node
625 evacuate, where we release locks earlier and thus allow higher
626 parallelism within the cluster
627 - Added watcher hooks, intended to allow the watcher to restart other
628 daemons (e.g. from the ganeti-nbma project), but they can be used of
629 course for any other purpose
630 - Added a compile-time disable for DRBD barriers, to increase
631 performance if the administrator trusts the power supply or the
632 storage system to not lose writes
633 - Added the option of using syslog for logging instead of, or in
634 addition to, Ganeti's own log files
635 - Removed boot restriction for paravirtual NICs for KVM, recent versions
636 can indeed boot from a paravirtual NIC
637 - Added a generic debug level for many operations; while this is not
638 used widely yet, it allows one to pass the debug value all the way to
640 - Enhanced the hooks environment for instance moves (failovers,
641 migrations) where the primary/secondary nodes changed during the
642 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
643 - Enhanced data validations for many user-supplied values; one important
644 item is the restrictions imposed on instance and node names, which
645 might reject some (invalid) host names
646 - Add a configure-time option to disable file-based storage, if it's not
647 needed; this allows greater security separation between the master
648 node and the other nodes from the point of view of the inter-node RPC
650 - Added user notification in interactive tools if job is waiting in the
651 job queue or trying to acquire locks
652 - Added log messages when a job is waiting for locks
653 - Added filtering by node tags in instance operations which admit
654 multiple instances (start, stop, reboot, reinstall)
655 - Added a new tool for cluster mergers, ``cluster-merge``
656 - Parameters from command line which are of the form ``a=b,c=d`` can now
657 use backslash escapes to pass in values which contain commas,
658 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
660 - For KVM, the instance name is the first parameter passed to KVM, so
661 that it's more visible in the process list
667 *(Released Tue, 2 Mar 2010)*
669 Ganeti 2.1 brings many improvements with it. Major changes:
671 - Added infrastructure to ease automated disk repairs
672 - Added new daemon to export configuration data in a cheaper way than
674 - Instance NICs can now be routed instead of being associated with a
676 - Improved job locking logic to reduce impact of jobs acquiring multiple
677 locks waiting for other long-running jobs
679 In-depth implementation details can be found in the Ganeti 2.1 design
685 - Added chroot hypervisor
686 - Added more options to xen-hvm hypervisor (``kernel_path`` and
688 - Added more options to xen-pvm hypervisor (``use_bootloader``,
689 ``bootloader_path`` and ``bootloader_args``)
690 - Added the ``use_localtime`` option for the xen-hvm and kvm
691 hypervisors, and the default value for this has changed to false (in
692 2.0 xen-hvm always enabled it)
693 - Added luxi call to submit multiple jobs in one go
694 - Added cluster initialization option to not modify ``/etc/hosts``
696 - Added network interface parameters
697 - Added dry run mode to some LUs
698 - Added RAPI resources:
700 - ``/2/instances/[instance_name]/info``
701 - ``/2/instances/[instance_name]/replace-disks``
702 - ``/2/nodes/[node_name]/evacuate``
703 - ``/2/nodes/[node_name]/migrate``
704 - ``/2/nodes/[node_name]/role``
705 - ``/2/nodes/[node_name]/storage``
706 - ``/2/nodes/[node_name]/storage/modify``
707 - ``/2/nodes/[node_name]/storage/repair``
709 - Added OpCodes to evacuate or migrate all instances on a node
710 - Added new command to list storage elements on nodes (``gnt-node
711 list-storage``) and modify them (``gnt-node modify-storage``)
712 - Added new ssconf files with master candidate IP address
713 (``ssconf_master_candidates_ips``), node primary IP address
714 (``ssconf_node_primary_ips``) and node secondary IP address
715 (``ssconf_node_secondary_ips``)
716 - Added ``ganeti-confd`` and a client library to query the Ganeti
717 configuration via UDP
718 - Added ability to run hooks after cluster initialization and before
720 - Added automatic mode for disk replace (``gnt-instance replace-disks
722 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
723 after catastrophic data-loss
724 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
726 - Added ``gnt-instance move`` command to move instances
727 - Added ``gnt-cluster watcher`` command to control watcher
728 - Added ``gnt-node powercycle`` command to powercycle nodes
729 - Added new job status field ``lock_status``
730 - Added parseable error codes to cluster verification (``gnt-cluster
731 verify --error-codes``) and made output less verbose (use
732 ``--verbose`` to restore previous behaviour)
733 - Added UUIDs to the main config entities (cluster, nodes, instances)
734 - Added support for OS variants
735 - Added support for hashed passwords in the Ganeti remote API users file
737 - Added option to specify maximum timeout on instance shutdown
738 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
739 - Added new helper script to start and stop Ganeti daemons
740 (``daemon-util``), with the intent to reduce the work necessary to
741 adjust Ganeti for non-Debian distributions and to start/stop daemons
743 - Added more unittests
744 - Fixed critical bug in ganeti-masterd startup
745 - Removed the configure-time ``kvm-migration-port`` parameter, this is
746 now customisable at the cluster level for both the KVM and Xen
747 hypervisors using the new ``migration_port`` parameter
748 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
749 reinstalling an instance
750 - Allowed ``@`` in tag names
751 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
752 - Many documentation updates
753 - Distribute hypervisor files on ``gnt-cluster redist-conf``
754 - ``gnt-instance reinstall`` can now reinstall multiple instances
755 - Updated many command line parameters
756 - Introduced new OS API version 15
757 - No longer support a default hypervisor
758 - Treat virtual LVs as inexistent
759 - Improved job locking logic to reduce lock contention
760 - Match instance and node names case insensitively
761 - Reimplemented bash completion script to be more complete
768 *(Released Thu, 4 Feb 2010)*
770 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
771 - Fix a string formatting bug
772 - Improve safety of the code in some error paths
773 - Improve data validation in the master of values returned from nodes
779 *(Released Thu, 17 Dec 2009)*
781 - Fix security issue due to missing validation of iallocator names; this
782 allows local and remote execution of arbitrary executables
783 - Fix failure of gnt-node list during instance removal
784 - Ship the RAPI documentation in the archive
790 *(Released Wed, 30 Sep 2009)*
792 - Fixed many wrong messages
793 - Fixed a few bugs related to the locking library
794 - Fixed MAC checking at instance creation time
795 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
796 - Fixed a few issues related to signal handling in both daemons and
798 - Fixed the example startup script provided
799 - Fixed insserv dependencies in the example startup script (patch from
801 - Fixed handling of drained nodes in the iallocator framework
802 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
804 - Fixed error related to invalid job IDs in job polling
805 - Fixed job/opcode persistence on unclean master shutdown
806 - Fixed handling of partial job processing after unclean master
808 - Fixed error reporting from LUs, previously all errors were converted
809 into execution errors
810 - Fixed error reporting from burnin
811 - Decreased significantly the memory usage of the job queue
812 - Optimised slightly multi-job submission
813 - Optimised slightly opcode loading
814 - Backported the multi-job submit framework from the development
815 branch; multi-instance start and stop should be faster
816 - Added script to clean archived jobs after 21 days; this will reduce
817 the size of the queue directory
818 - Added some extra checks in disk size tracking
819 - Added an example ethers hook script
820 - Added a cluster parameter that prevents Ganeti from modifying of
822 - Added more node information to RAPI responses
823 - Added a ``gnt-job watch`` command that allows following the ouput of a
825 - Added a bind-address option to ganeti-rapi
826 - Added more checks to the configuration verify
827 - Enhanced the burnin script such that some operations can be retried
829 - Converted instance reinstall to multi-instance model
835 *(Released Fri, 7 Aug 2009)*
837 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
838 to allow using the pre-2.0.2 behaviour in activation, if any existing
839 instances have mismatched disk sizes in the configuration
840 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
841 any configuration mismatches for disk sizes
842 - Added ``gnt-master cluste-failover --no-voting`` to allow master
843 failover to work on two-node clusters
844 - Fixed the ``--net`` option of ``gnt-backup import``, which was
846 - Fixed detection of OS script errors in ``gnt-backup export``
847 - Fixed exit code of ``gnt-backup export``
853 *(Released Fri, 17 Jul 2009)*
855 - Added experimental support for stripped logical volumes; this should
856 enhance performance but comes with a higher complexity in the block
857 device handling; stripping is only enabled when passing
858 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
859 affected even in the non-stripped mode
860 - Improved resiliency against transient failures at the end of DRBD
861 resyncs, and in general of DRBD resync checks
862 - Fixed a couple of issues with exports and snapshot errors
863 - Fixed a couple of issues in instance listing
864 - Added display of the disk size in ``gnt-instance info``
865 - Fixed checking for valid OSes in instance creation
866 - Fixed handling of the "vcpus" parameter in instance listing and in
867 general of invalid parameters
868 - Fixed http server library, and thus RAPI, to handle invalid
869 username/password combinations correctly; this means that now they
870 report unauthorized for queries too, not only for modifications,
871 allowing earlier detect of configuration problems
872 - Added a new "role" node list field, equivalent to the master/master
873 candidate/drained/offline flags combinations
874 - Fixed cluster modify and changes of candidate pool size
875 - Fixed cluster verify error messages for wrong files on regular nodes
876 - Fixed a couple of issues with node demotion from master candidate role
877 - Fixed node readd issues
878 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
879 - Added a new ``--no-voting`` option for masterfailover to fix failover
880 on two-nodes clusters when the former master node is unreachable
881 - Added instance reinstall over RAPI
887 *(Released Tue, 16 Jun 2009)*
889 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
890 allow re-adding the start in single-user option (regression from 1.2)
891 - the watcher writes the instance status to a file, to allow monitoring
892 to report the instance status (from the master) based on cached
893 results of the watcher's queries; while this can get stale if the
894 watcher is being locked due to other work on the cluster, this is
896 - the watcher now also restarts the node daemon and the rapi daemon if
898 - fixed the watcher to handle full and drained queue cases
899 - hooks export more instance data in the environment, which helps if
900 hook scripts need to take action based on the instance's properties
901 (no longer need to query back into ganeti)
902 - instance failovers when the instance is stopped do not check for free
903 RAM, so that failing over a stopped instance is possible in low memory
905 - rapi uses queries for tags instead of jobs (for less job traffic), and
906 for cluster tags it won't talk to masterd at all but read them from
908 - a couple of error handling fixes in RAPI
909 - drbd handling: improved the error handling of inconsistent disks after
910 resync to reduce the frequency of "there are some degraded disks for
911 this instance" messages
912 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
913 error handling path called a wrong function name)
919 *(Released Wed, 27 May 2009)*
921 - no changes from rc5
924 Version 2.0 release candidate 5
925 -------------------------------
927 *(Released Wed, 20 May 2009)*
929 - fix a couple of bugs (validation, argument checks)
930 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
931 - some small improvements to RAPI and IAllocator
932 - make watcher automatically start the master daemon if down
935 Version 2.0 release candidate 4
936 -------------------------------
938 *(Released Mon, 27 Apr 2009)*
940 - change the OS list to not require locks; this helps with big clusters
941 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
942 volume group is broken
943 - ``gnt-instance info``, without any arguments, doesn't run for all
944 instances anymore; either pass ``--all`` or pass the desired
945 instances; this helps against mistakes on big clusters where listing
946 the information for all instances takes a long time
947 - miscellaneous doc and man pages fixes
950 Version 2.0 release candidate 3
951 -------------------------------
953 *(Released Wed, 8 Apr 2009)*
955 - Change the internal locking model of some ``gnt-node`` commands, in
956 order to reduce contention (and blocking of master daemon) when
957 batching many creation/reinstall jobs
958 - Fixes to Xen soft reboot
959 - No longer build documentation at build time, instead distribute it in
960 the archive, in order to reduce the need for the whole docbook/rst
964 Version 2.0 release candidate 2
965 -------------------------------
967 *(Released Fri, 27 Mar 2009)*
969 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
970 - Fix watcher startup sequence, improves the behaviour of busy clusters
971 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
972 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
973 bind address and other places
974 - Some documentation fixes and updates
977 Version 2.0 release candidate 1
978 -------------------------------
980 *(Released Mon, 2 Mar 2009)*
982 - More documentation updates, now all docs should be more-or-less
984 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
986 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
988 - Fix an issue related to $libdir/run/ganeti and cluster creation
994 *(Released Thu, 19 Feb 2009)*
996 - Xen PVM and KVM have switched the default value for the instance root
997 disk to the first partition on the first drive, instead of the whole
998 drive; this means that the OS installation scripts must be changed
1000 - Man pages have been updated
1001 - RAPI has been switched by default to HTTPS, and the exported functions
1002 should all work correctly
1003 - RAPI v1 has been removed
1004 - Many improvements to the KVM hypervisor
1005 - Block device errors are now better reported
1006 - Many other bugfixes and small improvements
1012 *(Released Mon, 26 Jan 2009)*
1014 - Version 2 is a general rewrite of the code and therefore the
1015 differences are too many to list, see the design document for 2.0 in
1016 the ``doc/`` subdirectory for more details
1017 - In this beta version there is not yet a migration path from 1.2 (there
1018 will be one in the final 2.0 release)
1019 - A few significant changes are:
1021 - all commands are executed by a daemon (``ganeti-masterd``) and the
1022 various ``gnt-*`` commands are just front-ends to it
1023 - all the commands are entered into, and executed from a job queue,
1024 see the ``gnt-job(8)`` manpage
1025 - the RAPI daemon supports read-write operations, secured by basic
1026 HTTP authentication on top of HTTPS
1027 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1028 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1029 to migrate to DRBD 8 first while still running Ganeti 1.2)
1030 - DRBD devices are using statically allocated minor numbers, which
1031 will be assigned to existing instances during the migration process
1032 - there is support for both Xen PVM and Xen HVM instances running on
1034 - KVM virtualization is supported too
1035 - file-based storage has been implemented, which means that it is
1036 possible to run the cluster without LVM and DRBD storage, for
1037 example using a shared filesystem exported from shared storage (and
1038 still have live migration)
1044 *(Released Tue, 13 Jan 2009)*
1046 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1047 - Reuse the old instance mac address by default on instance import, if
1048 the instance name is the same.
1049 - Handle situations in which the node info rpc returns incomplete
1051 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1052 - Improved version of batcher:
1054 - state file support
1055 - instance mac address support
1056 - support for HVM clusters/instances
1058 - Add an option to show the number of cpu sockets and nodes in
1060 - Support OSes that handle more than one version of the OS api (but do
1061 not change the current API in any other way)
1062 - Fix ``gnt-node migrate``
1063 - ``gnt-debug`` man page
1064 - Fixes various more typos and small issues
1065 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1071 *(Released Wed, 24 Sep 2008)*
1073 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1074 type of disk exported to fully virtualized instances.
1075 - provide access to the serial console of HVM instances
1076 - instance auto_balance flag, set by default. If turned off it will
1077 avoid warnings on cluster verify if there is not enough memory to fail
1078 over an instance. in the future it will prevent automatically failing
1079 it over when we will support that.
1080 - batcher tool for instance creation, see ``tools/README.batcher``
1081 - ``gnt-instance reinstall --select-os`` to interactively select a new
1082 operating system when reinstalling an instance.
1083 - when changing the memory amount on instance modify a check has been
1084 added that the instance will be able to start. also warnings are
1085 emitted if the instance will not be able to fail over, if auto_balance
1087 - documentation fixes
1088 - sync fields between ``gnt-instance list/modify/add/import``
1089 - fix a race condition in drbd when the sync speed was set after giving
1090 the device a remote peer.
1096 *(Released Tue, 22 Jul 2008)*
1098 - note: the allowed size and number of tags per object were reduced
1099 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1100 - fixed twisted 8.x compatibility
1101 - fixed ``gnt-instance replace-disks`` with iallocator
1102 - add TCP keepalives on twisted connections to detect restarted nodes
1103 - disk increase support, see ``gnt-instance grow-disk``
1104 - implement bulk node/instance query for RAPI
1105 - add tags in node/instance listing (optional)
1106 - experimental migration (and live migration) support, read the man page
1107 for ``gnt-instance migrate``
1108 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1109 has some small improvements in handling its state file
1115 *(Released Fri, 13 Jun 2008)*
1117 - Experimental readonly, REST-based remote API implementation;
1118 automatically started on master node, TCP port 5080, if enabled by
1119 ``--enable-rapi`` parameter to configure script.
1120 - Instance allocator support. Add and import instance accept a
1121 ``--iallocator`` parameter, and call that instance allocator to decide
1122 which node to use for the instance. The iallocator document describes
1123 what's expected from an allocator script.
1124 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1125 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1126 node is lost there is still enough memory to fail over the instances
1128 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1129 ``gnt-cluster verify``, to check for site-specific compliance. All the
1130 hooks will run, and their output, if any, will be displayed. Any
1131 failing hook will make the verification return an error value.
1132 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1133 primary and secondary interfaces
1134 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1135 that is still declared as part of the cluster and has failed.
1136 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1137 specifying output fields, that just adds the chosen fields to the
1139 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1140 export from the filesystem.
1141 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1142 have been added. Using them you can enable/disable acpi and pae
1143 support, and specify a path for a cd image to be exported to the
1144 instance. These parameters as the name suggest only work on HVM
1146 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1147 PAE support will be set to the previously hardcoded values, but the
1148 (previously hardcoded) path to the CDROM ISO image will be unset and
1149 if required, needs to be set manually with ``gnt-instance modify``
1151 - The address to which an instance's VNC console is bound is now
1152 selectable per-instance, rather than being cluster wide. Of course
1153 this only applies to instances controlled via VNC, so currently just
1154 applies to HVM clusters.
1160 *(Released Mon, 18 Feb 2008)*
1162 - more tweaks to the disk activation code (especially helpful for DRBD)
1163 - change the default ``gnt-instance list`` output format, now there is
1164 one combined status field (see the manpage for the exact values this
1166 - some more fixes for the mac export to hooks change
1167 - make Ganeti not break with DRBD 8.2.x (which changed the version
1168 format in ``/proc/drbd``) (issue 24)
1169 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1170 template, allowing migration from DRBD0.7+MD to DRBD8
1176 *(Released Wed, 30 Jan 2008)*
1178 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1180 - add command aliases infrastructure and a few aliases
1181 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1182 man pages and the ``--help`` option of ``gnt-node
1183 list``/``gnt-instance list``
1184 - fix ``gnt-backup list`` with down nodes (issue 21)
1185 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1186 - fix the dist archive and add a check for including svn/git files in
1188 - some developer-related changes: improve the burnin and the QA suite,
1189 add an upload script for testing during development
1195 *(Released Wed, 16 Jan 2008)*
1197 - experimental HVM support, read the install document, section
1198 "Initializing the cluster"
1199 - allow for the PVM hypervisor per-instance kernel and initrd paths
1200 - add a new command ``gnt-cluster verify-disks`` which uses a new
1201 algorithm to improve the reconnection of the DRBD pairs if the device
1202 on the secondary node has gone away
1203 - make logical volume code auto-activate LVs at disk activation time
1204 - slightly improve the speed of activating disks
1205 - allow specification of the MAC address at instance creation time, and
1206 changing it later via ``gnt-instance modify``
1207 - fix handling of external commands that generate lots of output on
1209 - update documentation with regard to minimum version of DRBD8 supported
1215 *(Released Tue, 4 Dec 2007)*
1217 - Log the ``xm create`` output to the node daemon log on failure (to
1218 help diagnosing the error)
1219 - In debug mode, log all external commands output if failed to the logs
1220 - Change parsing of lvm commands to ignore stderr
1226 *(Released Wed, 28 Nov 2007)*
1228 - Another round of updates to the DRBD 8 code to deal with more failures
1229 in the replace secondary node operation
1230 - Some more logging of failures in disk operations (lvm, drbd)
1231 - A few documentation updates
1238 *(Released Tue, 13 Nov 2007)*
1240 - Change configuration file format from Python's Pickle to JSON.
1241 Upgrading is possible using the cfgupgrade utility.
1242 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1243 faster replace disks and is more stable (DRBD 8 has many improvements
1244 compared to DRBD 0.7)
1245 - Added command line tags support (see man pages for ``gnt-instance``,
1246 ``gnt-node``, ``gnt-cluster``)
1247 - Added instance rename support
1248 - Added multi-instance startup/shutdown
1249 - Added cluster rename support
1250 - Added ``gnt-node evacuate`` to simplify some node operations
1251 - Added instance reboot operation that can speedup reboot as compared to
1253 - Soften the requirement that hostnames are in FQDN format
1254 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1256 - Removed dependency on debian's patched fping that uses the
1257 non-standard ``-S`` option
1258 - Now the OS definitions are searched for in multiple, configurable
1259 paths (easier for distros to package)
1260 - Some changes to the hooks infrastructure (especially the new
1261 post-configuration update hook)
1262 - Other small bugfixes
1264 .. vim: set textwidth=72 syntax=rst :