9 - Moved ``rapi_users`` file into separate directory, now named
10 ``…/ganeti/rapi/users``
16 *(Released Fri, 19 Nov 2010)*
18 A number of bugfixes and documentation updates:
20 - Update ganeti-os-interface documentation
21 - Fixed a bug related to duplicate MACs or similar items which should be
23 - Fix breakage in OS state modify
24 - Reinstall instance: disallow offline secondaries (fixes bug related to
25 OS changing but reinstall failing)
26 - plus all the other fixes between 2.2.1 and 2.2.2
32 *(Released Tue, 2 Nov 2010)*
34 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
35 - Fixed master failover race with watcher
36 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
38 - Accept previously stopped instance for export with instance removal
39 - Simplify and extend the environment variables for instance OS scripts
40 - Added new node flags, ``master_capable`` and ``vm_capable``
41 - Added optional instance disk wiping prior during allocation. This is a
42 cluster-wide option and can be set/modified using
43 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
44 - Added IPv6 support, see :doc:`design document <design-2.3>` and
46 - Added a new watcher option (``--ignore-pause``)
47 - Added option to ignore offline node on instance start/stop
48 (``--ignore-offline``)
49 - Allow overriding OS parameters with ``gnt-instance reinstall``
50 - Added ability to change node's secondary IP address using ``gnt-node
52 - Implemented privilege separation for all daemons except
53 ``ganeti-noded``, see ``configure`` options
54 - Complain if an instance's disk is marked faulty in ``gnt-cluster
56 - Implemented job priorities (see ``ganeti(7)`` manpage)
57 - Ignore failures while shutting down instances during failover from
59 - Exit daemon's bootstrap process only once daemon is ready
60 - Export more information via ``LUQueryInstances``/remote API
61 - Improved documentation, QA and unittests
62 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
63 restart if the file was created or changed
64 - Added LUXI protocol version sent with each request and response,
65 allowing detection of server/client mismatches
66 - Moved the Python scripts among gnt-* and ganeti-* into modules
67 - Moved all code related to setting up SSH to an external script,
69 - Infrastructure changes for node group support in future versions
75 *(Released Fri, 19 Nov 2010)*
77 A few small bugs fixed, and some improvements to the build system:
79 - Fix documentation regarding conversion to drbd
80 - Fix validation of parameters in cluster modify (``gnt-cluster modify
82 - Fix error handling in node modify with multiple changes
83 - Allow remote imports without checked names
89 *(Released Tue, 19 Oct 2010)*
91 - Disable SSL session ID cache in RPC client
97 *(Released Thu, 14 Oct 2010)*
99 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
100 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
102 - Fix problems with interaction between Curl and Python's HTTP server,
103 resulting in increased speed in many RPC calls
104 - Improve our release script to prevent breakage with older aclocal and
111 *(Released Thu, 7 Oct 2010)*
113 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
114 value retrieved from master
115 - Added support for blacklisted or hidden OS definitions
116 - Added simple lock monitor (accessible via (``gnt-debug locks``)
117 - Added support for -mem-path in KVM hypervisor abstraction layer
118 - Allow overriding instance parameters in tool for inter-cluster
119 instance moves (``tools/move-instance``)
120 - Improved opcode summaries (e.g. in ``gnt-job list``)
121 - Improve consistency of OS listing by sorting it
122 - Documentation updates
128 *(Released Fri, 8 Oct 2010)*
130 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
136 *(Released Mon, 4 Oct 2010)*
138 - Fixed regression in ``gnt-instance rename``
144 *(Released Wed, 22 Sep 2010)*
146 - Fixed OS_VARIANT variable for OS scripts
147 - Fixed cluster tag operations via RAPI
148 - Made ``setup-ssh`` exit with non-zero code if an error occurred
149 - Disabled RAPI CA checks in watcher
155 *(Released Mon, 23 Aug 2010)*
157 - Support DRBD versions of the format "a.b.c.d"
159 - Re-introduce support for usage from multiple threads in RAPI client
160 - Instance renames and modify via RAPI
161 - Work around race condition between processing and archival in job
163 - Mark opcodes following failed one as failed, too
164 - Job field ``lock_status`` was removed due to difficulties making it
165 work with the changed job queue in Ganeti 2.2; a better way to monitor
166 locks is expected for a later 2.2.x release
167 - Fixed dry-run behaviour with many commands
168 - Support ``ssh-agent`` again when adding nodes
169 - Many additional bugfixes
175 *(Released Fri, 30 Jul 2010)*
177 Important change: the internal RPC mechanism between Ganeti nodes has
178 changed from using a home-grown http library (based on the Python base
179 libraries) to use the PycURL library. This requires that PycURL is
180 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
181 against GnuTLS by default. cURL's support for GnuTLS had known issues
182 before cURL 7.21.0 and we recommend using the latest cURL release or
183 linking against OpenSSL. Most other distributions already link PycURL
184 and cURL against OpenSSL. The command::
186 python -c 'import pycurl; print pycurl.version'
188 can be used to determine the libraries PycURL and cURL are linked
191 Other significant changes:
193 - Rewrote much of the internals of the job queue, in order to achieve
194 better parallelism; this decouples job query operations from the job
195 processing, and it should allow much nicer behaviour of the master
196 daemon under load, and it also has uncovered some long-standing bugs
197 related to the job serialisation (now fixed)
198 - Added a default iallocator setting to the cluster parameters,
199 eliminating the need to always pass nodes or an iallocator for
200 operations that require selection of new node(s)
201 - Added experimental support for the LXC virtualization method
202 - Added support for OS parameters, which allows the installation of
203 instances to pass parameter to OS scripts in order to customise the
205 - Added a hypervisor parameter controlling the migration type (live or
206 non-live), since hypervisors have various levels of reliability; this
207 has renamed the 'live' parameter to 'mode'
208 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
209 logical volumes, meaning that cluster verify will ignore them and not
210 flag their presence as errors
211 - The watcher will now reset the error count for failed instances after
212 8 hours, thus allowing self-healing if the problem that caused the
213 instances to be down/fail to start has cleared in the meantime
214 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
215 check for, and warn, if the drbd module parameter ``usermode_helper``
216 is not consistent with the cluster-wide setting; this is needed to
217 make diagnose easier of failed drbd creations
218 - Started adding base IPv6 support, but this is not yet
219 enabled/available for use
220 - Rename operations (cluster, instance) will now return the new name,
221 which is especially useful if a short name was passed in
222 - Added support for instance migration in RAPI
223 - Added a tool to pre-configure nodes for the SSH setup, before joining
224 them to the cluster; this will allow in the future a simplified model
225 for node joining (but not yet fully enabled in 2.2); this needs the
226 paramiko python library
227 - Fixed handling of name-resolving errors
228 - Fixed consistency of job results on the error path
229 - Fixed master-failover race condition when executed multiple times in
231 - Fixed many bugs related to the job queue (mostly introduced during the
232 2.2 development cycle, so not all are impacting 2.1)
233 - Fixed instance migration with missing disk symlinks
234 - Fixed handling of unknown jobs in ``gnt-job archive``
235 - And many other small fixes/improvements
239 - Enhanced both the unittest and the QA coverage
240 - Switched the opcode validation to a generic model, and extended the
241 validation to all opcode parameters
242 - Changed more parts of the code that write shell scripts to use the
244 - Switched the master daemon to use the asyncore library for the Luxi
251 *(Released Thu, 17 Jun 2010)*
253 - Added tool (``move-instance``) and infrastructure to move instances
254 between separate clusters (see :doc:`separate documentation
255 <move-instance>` and :doc:`design document <design-2.2>`)
256 - Added per-request RPC timeout
257 - RAPI now requires a Content-Type header for requests with a body (e.g.
258 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
259 :rfc:`2616` (HTTP/1.1), section 7.2.1)
260 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
262 - Implemented initial support for running Ganeti daemons as separate
263 users, see configure-time flags ``--with-user-prefix`` and
264 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
266 - Instances can be removed after export (``gnt-backup export
268 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
269 key (instead of 1024 bit)
270 - Added new cluster configuration file for cluster domain secret
271 - Import/export now use SSL instead of SSH
272 - Added support for showing estimated time when exporting an instance,
273 see the ``ganeti-os-interface(7)`` manpage and look for
280 *(Released Tue, 16 Nov 2010)*
282 Some more bugfixes. Unless critical bugs occur, this will be the last
285 - Fix case of MAC special-values
286 - Fix mac checker regex
287 - backend: Fix typo causing “out of range” error
288 - Add missing --units in gnt-instance list man page
294 *(Released Tue, 24 Aug 2010)*
297 - Don't ignore secondary node silently on non-mirrored disk templates
299 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
300 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
301 - Properly document the usb_mouse parameter
302 - Fix path in ganeti-rapi(8) (issue 116)
303 - Adjust error message when the ganeti user's .ssh directory is
305 - Add same-node-check when changing the disk template to drbd
311 *(Released Fri, 16 Jul 2010)*
314 - Add an option to only select some reboot types during qa/burnin.
315 (on some hypervisors consequent reboots are not supported)
316 - Fix infrequent race condition in master failover. Sometimes the old
317 master ip address would be still detected as up for a short time
318 after it was removed, causing failover to fail.
319 - Decrease mlockall warnings when the ctypes module is missing. On
320 Python 2.4 we support running even if no ctypes module is installed,
321 but we were too verbose about this issue.
322 - Fix building on old distributions, on which man doesn't have a
324 - Fix RAPI not to ignore the MAC address on instance creation
325 - Implement the old instance creation format in the RAPI client.
331 *(Released Thu, 01 Jul 2010)*
333 A small bugfix release:
334 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
335 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
336 a non-existing option
337 - Fix instance create: the --force-variant option was ignored
338 - Improve pylint 0.21 compatibility and warnings with Python 2.6
339 - Fix modify node storage with non-FQDN arguments
340 - Fix RAPI client to authenticate under Python 2.6 when used
341 for more than 5 requests needing authentication
342 - Fix gnt-instance modify -t (storage) giving a wrong error message
343 when converting a non-shutdown drbd instance to plain
349 *(Released Fri, 18 Jun 2010)*
351 A small bugfix release:
353 - Fix live migration of KVM instances started with older Ganeti
354 versions which had fewer hypervisor parameters
355 - Fix gnt-instance grow-disk on down instances
356 - Fix an error-reporting bug during instance migration
357 - Better checking of the ``--net`` and ``--disk`` values, to avoid
358 silently ignoring broken ones
359 - Fix an RPC error reporting bug affecting, for example, RAPI client
361 - Fix bug triggered by different API version os-es on different nodes
362 - Fix a bug in instance startup with custom hvparams: OS level
363 parameters would fail to be applied.
364 - Fix the RAPI client under Python 2.6 (but more work is needed to
365 make it work completely well with OpenSSL)
366 - Fix handling of errors when resolving names from DNS
372 *(Released Thu, 3 Jun 2010)*
374 A medium sized development cycle. Some new features, and some
375 fixes/small improvements/cleanups.
380 The node deamon now tries to mlock itself into memory, unless the
381 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
382 its logs, and falls back to console logging. This allows emergency
383 features such as ``gnt-node powercycle`` to work even in the event of a
384 broken node disk (tested offlining the disk hosting the node's
385 filesystem and dropping its memory caches; don't try this at home)
387 KVM: add vhost-net acceleration support. It can be tested with a new
388 enough version of the kernel and of qemu-kvm.
390 KVM: Add instance chrooting feature. If you use privilege dropping for
391 your VMs you can also now force them to chroot to an empty directory,
392 before starting the emulated guest.
394 KVM: Add maximum migration bandwith and maximum downtime tweaking
395 support (requires a new-enough version of qemu-kvm).
397 Cluster verify will now warn if the master node doesn't have the master
400 Add a new (incompatible) instance creation request format to RAPI which
401 supports all parameters (previously only a subset was supported, and it
402 wasn't possible to extend the old format to accomodate all the new
403 features. The old format is still supported, and a client can check for
404 this feature, before using it, by checking for its presence in the
405 ``features`` RAPI resource.
407 Now with ancient latin support. Try it passing the ``--roman`` option to
408 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
409 (requires the python-roman module to be installed, in order to work).
414 As usual many internal code refactorings, documentation updates, and
417 - Lots of improvements and cleanups to the experimental Remote API
418 (RAPI) client library.
419 - A new unit test suite for the core daemon libraries.
420 - A fix to creating missing directories makes sure the umask is not
421 applied anymore. This enforces the same directory permissions
423 - Better handling terminating daemons with ctrl+c (used when running
424 them in debugging mode).
425 - Fix a race condition in live migrating a KVM instance, when stat()
426 on the old proc status file returned EINVAL, which is an unexpected
428 - Fixed manpage checking with newer man and utf-8 charachters. But now
429 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
435 *(Released Fri, 7 May 2010)*
437 Fix a bug which prevented untagged KVM instances from starting.
443 *(Released Fri, 7 May 2010)*
445 Another release with a long development cycle, during which many
446 different features were added.
451 The KVM hypervisor now can run the individual instances as non-root, to
452 reduce the impact of a VM being hijacked due to bugs in the
453 hypervisor. It is possible to run all instances as a single (non-root)
454 user, to manually specify a user for each instance, or to dynamically
455 allocate a user out of a cluster-wide pool to each instance, with the
456 guarantee that no two instances will run under the same user ID on any
459 An experimental RAPI client library, that can be used standalone
460 (without the other Ganeti libraries), is provided in the source tree as
461 ``lib/rapi/client.py``. Note this client might change its interface in
462 the future, as we iterate on its capabilities.
464 A new command, ``gnt-cluster renew-crypto`` has been added to easily
465 replace the cluster's certificates and crypto keys. This might help in
466 case they have been compromised, or have simply expired.
468 A new disk option for instance creation has been added that allows one
469 to "adopt" currently existing logical volumes, with data
470 preservation. This should allow easier migration to Ganeti from
471 unmanaged (or managed via other software) instances.
473 Another disk improvement is the possibility to convert between redundant
474 (DRBD) and plain (LVM) disk configuration for an instance. This should
475 allow better scalability (starting with one node and growing the
476 cluster, or shrinking a two-node cluster to one node).
478 A new feature that could help with automated node failovers has been
479 implemented: if a node sees itself as offline (by querying the master
480 candidates), it will try to shutdown (hard) all instances and any active
481 DRBD devices. This reduces the risk of duplicate instances if an
482 external script automatically failovers the instances on such nodes. To
483 enable this, the cluster parameter ``maintain_node_health`` should be
484 enabled; in the future this option (per the name) will enable other
485 automatic maintenance features.
487 Instance export/import now will reuse the original instance
488 specifications for all parameters; that means exporting an instance,
489 deleting it and the importing it back should give an almost identical
490 instance. Note that the default import behaviour has changed from
491 before, where it created only one NIC; now it recreates the original
494 Cluster verify has added a few new checks: SSL certificates validity,
495 /etc/hosts consistency across the cluster, etc.
500 As usual, many internal changes were done, documentation fixes,
503 - Fixed cluster initialization with disabled cluster storage (regression
505 - File-based storage supports growing the disks
506 - Fixed behaviour of node role changes
507 - Fixed cluster verify for some corner cases, plus a general rewrite of
508 cluster verify to allow future extension with more checks
509 - Fixed log spamming by watcher and node daemon (regression introduced
511 - Fixed possible validation issues when changing the list of enabled
513 - Fixed cleanup of /etc/hosts during node removal
514 - Fixed RAPI response for invalid methods
515 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
516 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
517 from ide disks, etc.)
518 - Allow OS changes without re-installation (to record a changed OS
519 outside of Ganeti, or to allow OS renames)
520 - Allow instance creation without OS installation (useful for example if
521 the OS will be installed manually, or restored from a backup not in
523 - Implemented option to make cluster ``copyfile`` use the replication
525 - Added list of enabled hypervisors to ssconf (possibly useful for
527 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
529 - A partial form of node re-IP is possible via node readd, which now
530 allows changed node primary IP
531 - Command line utilities now show an informational message if the job is
533 - The logs of the master daemon now show the PID/UID/GID of the
540 *(Released Fri, 12 Mar 2010)*
542 During the 2.1.0 long release candidate cycle, a lot of improvements and
543 changes have accumulated with were released later as 2.1.1.
548 The node evacuate command (``gnt-node evacuate``) was significantly
549 rewritten, and as such the IAllocator protocol was changed - a new
550 request type has been added. This unfortunate change during a stable
551 series is designed to improve performance of node evacuations; on
552 clusters with more than about five nodes and which are well-balanced,
553 evacuation should proceed in parallel for all instances of the node
554 being evacuated. As such, any existing IAllocator scripts need to be
555 updated, otherwise the above command will fail due to the unknown
556 request. The provided "dumb" allocator has not been updated; but the
557 ganeti-htools package supports the new protocol since version 0.2.4.
559 Another important change is increased validation of node and instance
560 names. This might create problems in special cases, if invalid host
561 names are being used.
563 Also, a new layer of hypervisor parameters has been added, that sits at
564 OS level between the cluster defaults and the instance ones. This allows
565 customisation of virtualization parameters depending on the installed
566 OS. For example instances with OS 'X' may have a different KVM kernel
567 (or any other parameter) than the cluster defaults. This is intended to
568 help managing a multiple OSes on the same cluster, without manual
569 modification of each instance's parameters.
571 A tool for merging clusters, ``cluster-merge``, has been added in the
577 - Improved the int/float conversions that should make the code more
578 robust in face of errors from the node daemons
579 - Fixed the remove node code in case of internal configuration errors
580 - Fixed the node daemon behaviour in face of inconsistent queue
581 directory (e.g. read-only file-system where we can't open the files
583 - Fixed the behaviour of gnt-node modify for master candidate demotion;
584 now it either aborts cleanly or, if given the new “auto_promote”
585 parameter, will automatically promote other nodes as needed
586 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
587 completely prevent Ganeti from working
588 - Fixed bug for instance export when not all disks were successfully
590 - Fixed behaviour of node add when the new node is slow in starting up
592 - Fixed handling of signals in the LUXI client, which should improve
593 behaviour of command-line scripts
594 - Added checks for invalid node/instance names in the configuration (now
595 flagged during cluster verify)
596 - Fixed watcher behaviour for disk activation errors
597 - Fixed two potentially endless loops in http library, which led to the
598 RAPI daemon hanging and consuming 100% CPU in some cases
599 - Fixed bug in RAPI daemon related to hashed passwords
600 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
602 - Enhanced compatibility with non-Debian OSes, but not using absolute
603 path in some commands and allowing customisation of the ssh
604 configuration directory
605 - Fixed possible future issue with new Python versions by abiding to the
606 proper use of ``__slots__`` attribute on classes
607 - Added checks that should prevent directory traversal attacks
608 - Many documentation fixes based on feedback from users
613 - Added an “early_release” more for instance replace disks and node
614 evacuate, where we release locks earlier and thus allow higher
615 parallelism within the cluster
616 - Added watcher hooks, intended to allow the watcher to restart other
617 daemons (e.g. from the ganeti-nbma project), but they can be used of
618 course for any other purpose
619 - Added a compile-time disable for DRBD barriers, to increase
620 performance if the administrator trusts the power supply or the
621 storage system to not lose writes
622 - Added the option of using syslog for logging instead of, or in
623 addition to, Ganeti's own log files
624 - Removed boot restriction for paravirtual NICs for KVM, recent versions
625 can indeed boot from a paravirtual NIC
626 - Added a generic debug level for many operations; while this is not
627 used widely yet, it allows one to pass the debug value all the way to
629 - Enhanced the hooks environment for instance moves (failovers,
630 migrations) where the primary/secondary nodes changed during the
631 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
632 - Enhanced data validations for many user-supplied values; one important
633 item is the restrictions imposed on instance and node names, which
634 might reject some (invalid) host names
635 - Add a configure-time option to disable file-based storage, if it's not
636 needed; this allows greater security separation between the master
637 node and the other nodes from the point of view of the inter-node RPC
639 - Added user notification in interactive tools if job is waiting in the
640 job queue or trying to acquire locks
641 - Added log messages when a job is waiting for locks
642 - Added filtering by node tags in instance operations which admit
643 multiple instances (start, stop, reboot, reinstall)
644 - Added a new tool for cluster mergers, ``cluster-merge``
645 - Parameters from command line which are of the form ``a=b,c=d`` can now
646 use backslash escapes to pass in values which contain commas,
647 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
649 - For KVM, the instance name is the first parameter passed to KVM, so
650 that it's more visible in the process list
656 *(Released Tue, 2 Mar 2010)*
658 Ganeti 2.1 brings many improvements with it. Major changes:
660 - Added infrastructure to ease automated disk repairs
661 - Added new daemon to export configuration data in a cheaper way than
663 - Instance NICs can now be routed instead of being associated with a
665 - Improved job locking logic to reduce impact of jobs acquiring multiple
666 locks waiting for other long-running jobs
668 In-depth implementation details can be found in the Ganeti 2.1 design
674 - Added chroot hypervisor
675 - Added more options to xen-hvm hypervisor (``kernel_path`` and
677 - Added more options to xen-pvm hypervisor (``use_bootloader``,
678 ``bootloader_path`` and ``bootloader_args``)
679 - Added the ``use_localtime`` option for the xen-hvm and kvm
680 hypervisors, and the default value for this has changed to false (in
681 2.0 xen-hvm always enabled it)
682 - Added luxi call to submit multiple jobs in one go
683 - Added cluster initialization option to not modify ``/etc/hosts``
685 - Added network interface parameters
686 - Added dry run mode to some LUs
687 - Added RAPI resources:
689 - ``/2/instances/[instance_name]/info``
690 - ``/2/instances/[instance_name]/replace-disks``
691 - ``/2/nodes/[node_name]/evacuate``
692 - ``/2/nodes/[node_name]/migrate``
693 - ``/2/nodes/[node_name]/role``
694 - ``/2/nodes/[node_name]/storage``
695 - ``/2/nodes/[node_name]/storage/modify``
696 - ``/2/nodes/[node_name]/storage/repair``
698 - Added OpCodes to evacuate or migrate all instances on a node
699 - Added new command to list storage elements on nodes (``gnt-node
700 list-storage``) and modify them (``gnt-node modify-storage``)
701 - Added new ssconf files with master candidate IP address
702 (``ssconf_master_candidates_ips``), node primary IP address
703 (``ssconf_node_primary_ips``) and node secondary IP address
704 (``ssconf_node_secondary_ips``)
705 - Added ``ganeti-confd`` and a client library to query the Ganeti
706 configuration via UDP
707 - Added ability to run hooks after cluster initialization and before
709 - Added automatic mode for disk replace (``gnt-instance replace-disks
711 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
712 after catastrophic data-loss
713 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
715 - Added ``gnt-instance move`` command to move instances
716 - Added ``gnt-cluster watcher`` command to control watcher
717 - Added ``gnt-node powercycle`` command to powercycle nodes
718 - Added new job status field ``lock_status``
719 - Added parseable error codes to cluster verification (``gnt-cluster
720 verify --error-codes``) and made output less verbose (use
721 ``--verbose`` to restore previous behaviour)
722 - Added UUIDs to the main config entities (cluster, nodes, instances)
723 - Added support for OS variants
724 - Added support for hashed passwords in the Ganeti remote API users file
726 - Added option to specify maximum timeout on instance shutdown
727 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
728 - Added new helper script to start and stop Ganeti daemons
729 (``daemon-util``), with the intent to reduce the work necessary to
730 adjust Ganeti for non-Debian distributions and to start/stop daemons
732 - Added more unittests
733 - Fixed critical bug in ganeti-masterd startup
734 - Removed the configure-time ``kvm-migration-port`` parameter, this is
735 now customisable at the cluster level for both the KVM and Xen
736 hypervisors using the new ``migration_port`` parameter
737 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
738 reinstalling an instance
739 - Allowed ``@`` in tag names
740 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
741 - Many documentation updates
742 - Distribute hypervisor files on ``gnt-cluster redist-conf``
743 - ``gnt-instance reinstall`` can now reinstall multiple instances
744 - Updated many command line parameters
745 - Introduced new OS API version 15
746 - No longer support a default hypervisor
747 - Treat virtual LVs as inexistent
748 - Improved job locking logic to reduce lock contention
749 - Match instance and node names case insensitively
750 - Reimplemented bash completion script to be more complete
757 *(Released Thu, 4 Feb 2010)*
759 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
760 - Fix a string formatting bug
761 - Improve safety of the code in some error paths
762 - Improve data validation in the master of values returned from nodes
768 *(Released Thu, 17 Dec 2009)*
770 - Fix security issue due to missing validation of iallocator names; this
771 allows local and remote execution of arbitrary executables
772 - Fix failure of gnt-node list during instance removal
773 - Ship the RAPI documentation in the archive
779 *(Released Wed, 30 Sep 2009)*
781 - Fixed many wrong messages
782 - Fixed a few bugs related to the locking library
783 - Fixed MAC checking at instance creation time
784 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
785 - Fixed a few issues related to signal handling in both daemons and
787 - Fixed the example startup script provided
788 - Fixed insserv dependencies in the example startup script (patch from
790 - Fixed handling of drained nodes in the iallocator framework
791 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
793 - Fixed error related to invalid job IDs in job polling
794 - Fixed job/opcode persistence on unclean master shutdown
795 - Fixed handling of partial job processing after unclean master
797 - Fixed error reporting from LUs, previously all errors were converted
798 into execution errors
799 - Fixed error reporting from burnin
800 - Decreased significantly the memory usage of the job queue
801 - Optimised slightly multi-job submission
802 - Optimised slightly opcode loading
803 - Backported the multi-job submit framework from the development
804 branch; multi-instance start and stop should be faster
805 - Added script to clean archived jobs after 21 days; this will reduce
806 the size of the queue directory
807 - Added some extra checks in disk size tracking
808 - Added an example ethers hook script
809 - Added a cluster parameter that prevents Ganeti from modifying of
811 - Added more node information to RAPI responses
812 - Added a ``gnt-job watch`` command that allows following the ouput of a
814 - Added a bind-address option to ganeti-rapi
815 - Added more checks to the configuration verify
816 - Enhanced the burnin script such that some operations can be retried
818 - Converted instance reinstall to multi-instance model
824 *(Released Fri, 7 Aug 2009)*
826 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
827 to allow using the pre-2.0.2 behaviour in activation, if any existing
828 instances have mismatched disk sizes in the configuration
829 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
830 any configuration mismatches for disk sizes
831 - Added ``gnt-master cluste-failover --no-voting`` to allow master
832 failover to work on two-node clusters
833 - Fixed the ``--net`` option of ``gnt-backup import``, which was
835 - Fixed detection of OS script errors in ``gnt-backup export``
836 - Fixed exit code of ``gnt-backup export``
842 *(Released Fri, 17 Jul 2009)*
844 - Added experimental support for stripped logical volumes; this should
845 enhance performance but comes with a higher complexity in the block
846 device handling; stripping is only enabled when passing
847 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
848 affected even in the non-stripped mode
849 - Improved resiliency against transient failures at the end of DRBD
850 resyncs, and in general of DRBD resync checks
851 - Fixed a couple of issues with exports and snapshot errors
852 - Fixed a couple of issues in instance listing
853 - Added display of the disk size in ``gnt-instance info``
854 - Fixed checking for valid OSes in instance creation
855 - Fixed handling of the "vcpus" parameter in instance listing and in
856 general of invalid parameters
857 - Fixed http server library, and thus RAPI, to handle invalid
858 username/password combinations correctly; this means that now they
859 report unauthorized for queries too, not only for modifications,
860 allowing earlier detect of configuration problems
861 - Added a new "role" node list field, equivalent to the master/master
862 candidate/drained/offline flags combinations
863 - Fixed cluster modify and changes of candidate pool size
864 - Fixed cluster verify error messages for wrong files on regular nodes
865 - Fixed a couple of issues with node demotion from master candidate role
866 - Fixed node readd issues
867 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
868 - Added a new ``--no-voting`` option for masterfailover to fix failover
869 on two-nodes clusters when the former master node is unreachable
870 - Added instance reinstall over RAPI
876 *(Released Tue, 16 Jun 2009)*
878 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
879 allow re-adding the start in single-user option (regression from 1.2)
880 - the watcher writes the instance status to a file, to allow monitoring
881 to report the instance status (from the master) based on cached
882 results of the watcher's queries; while this can get stale if the
883 watcher is being locked due to other work on the cluster, this is
885 - the watcher now also restarts the node daemon and the rapi daemon if
887 - fixed the watcher to handle full and drained queue cases
888 - hooks export more instance data in the environment, which helps if
889 hook scripts need to take action based on the instance's properties
890 (no longer need to query back into ganeti)
891 - instance failovers when the instance is stopped do not check for free
892 RAM, so that failing over a stopped instance is possible in low memory
894 - rapi uses queries for tags instead of jobs (for less job traffic), and
895 for cluster tags it won't talk to masterd at all but read them from
897 - a couple of error handling fixes in RAPI
898 - drbd handling: improved the error handling of inconsistent disks after
899 resync to reduce the frequency of "there are some degraded disks for
900 this instance" messages
901 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
902 error handling path called a wrong function name)
908 *(Released Wed, 27 May 2009)*
910 - no changes from rc5
913 Version 2.0 release candidate 5
914 -------------------------------
916 *(Released Wed, 20 May 2009)*
918 - fix a couple of bugs (validation, argument checks)
919 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
920 - some small improvements to RAPI and IAllocator
921 - make watcher automatically start the master daemon if down
924 Version 2.0 release candidate 4
925 -------------------------------
927 *(Released Mon, 27 Apr 2009)*
929 - change the OS list to not require locks; this helps with big clusters
930 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
931 volume group is broken
932 - ``gnt-instance info``, without any arguments, doesn't run for all
933 instances anymore; either pass ``--all`` or pass the desired
934 instances; this helps against mistakes on big clusters where listing
935 the information for all instances takes a long time
936 - miscellaneous doc and man pages fixes
939 Version 2.0 release candidate 3
940 -------------------------------
942 *(Released Wed, 8 Apr 2009)*
944 - Change the internal locking model of some ``gnt-node`` commands, in
945 order to reduce contention (and blocking of master daemon) when
946 batching many creation/reinstall jobs
947 - Fixes to Xen soft reboot
948 - No longer build documentation at build time, instead distribute it in
949 the archive, in order to reduce the need for the whole docbook/rst
953 Version 2.0 release candidate 2
954 -------------------------------
956 *(Released Fri, 27 Mar 2009)*
958 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
959 - Fix watcher startup sequence, improves the behaviour of busy clusters
960 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
961 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
962 bind address and other places
963 - Some documentation fixes and updates
966 Version 2.0 release candidate 1
967 -------------------------------
969 *(Released Mon, 2 Mar 2009)*
971 - More documentation updates, now all docs should be more-or-less
973 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
975 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
977 - Fix an issue related to $libdir/run/ganeti and cluster creation
983 *(Released Thu, 19 Feb 2009)*
985 - Xen PVM and KVM have switched the default value for the instance root
986 disk to the first partition on the first drive, instead of the whole
987 drive; this means that the OS installation scripts must be changed
989 - Man pages have been updated
990 - RAPI has been switched by default to HTTPS, and the exported functions
991 should all work correctly
992 - RAPI v1 has been removed
993 - Many improvements to the KVM hypervisor
994 - Block device errors are now better reported
995 - Many other bugfixes and small improvements
1001 *(Released Mon, 26 Jan 2009)*
1003 - Version 2 is a general rewrite of the code and therefore the
1004 differences are too many to list, see the design document for 2.0 in
1005 the ``doc/`` subdirectory for more details
1006 - In this beta version there is not yet a migration path from 1.2 (there
1007 will be one in the final 2.0 release)
1008 - A few significant changes are:
1010 - all commands are executed by a daemon (``ganeti-masterd``) and the
1011 various ``gnt-*`` commands are just front-ends to it
1012 - all the commands are entered into, and executed from a job queue,
1013 see the ``gnt-job(8)`` manpage
1014 - the RAPI daemon supports read-write operations, secured by basic
1015 HTTP authentication on top of HTTPS
1016 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1017 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1018 to migrate to DRBD 8 first while still running Ganeti 1.2)
1019 - DRBD devices are using statically allocated minor numbers, which
1020 will be assigned to existing instances during the migration process
1021 - there is support for both Xen PVM and Xen HVM instances running on
1023 - KVM virtualization is supported too
1024 - file-based storage has been implemented, which means that it is
1025 possible to run the cluster without LVM and DRBD storage, for
1026 example using a shared filesystem exported from shared storage (and
1027 still have live migration)
1033 *(Released Tue, 13 Jan 2009)*
1035 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1036 - Reuse the old instance mac address by default on instance import, if
1037 the instance name is the same.
1038 - Handle situations in which the node info rpc returns incomplete
1040 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1041 - Improved version of batcher:
1043 - state file support
1044 - instance mac address support
1045 - support for HVM clusters/instances
1047 - Add an option to show the number of cpu sockets and nodes in
1049 - Support OSes that handle more than one version of the OS api (but do
1050 not change the current API in any other way)
1051 - Fix ``gnt-node migrate``
1052 - ``gnt-debug`` man page
1053 - Fixes various more typos and small issues
1054 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1060 *(Released Wed, 24 Sep 2008)*
1062 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1063 type of disk exported to fully virtualized instances.
1064 - provide access to the serial console of HVM instances
1065 - instance auto_balance flag, set by default. If turned off it will
1066 avoid warnings on cluster verify if there is not enough memory to fail
1067 over an instance. in the future it will prevent automatically failing
1068 it over when we will support that.
1069 - batcher tool for instance creation, see ``tools/README.batcher``
1070 - ``gnt-instance reinstall --select-os`` to interactively select a new
1071 operating system when reinstalling an instance.
1072 - when changing the memory amount on instance modify a check has been
1073 added that the instance will be able to start. also warnings are
1074 emitted if the instance will not be able to fail over, if auto_balance
1076 - documentation fixes
1077 - sync fields between ``gnt-instance list/modify/add/import``
1078 - fix a race condition in drbd when the sync speed was set after giving
1079 the device a remote peer.
1085 *(Released Tue, 22 Jul 2008)*
1087 - note: the allowed size and number of tags per object were reduced
1088 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1089 - fixed twisted 8.x compatibility
1090 - fixed ``gnt-instance replace-disks`` with iallocator
1091 - add TCP keepalives on twisted connections to detect restarted nodes
1092 - disk increase support, see ``gnt-instance grow-disk``
1093 - implement bulk node/instance query for RAPI
1094 - add tags in node/instance listing (optional)
1095 - experimental migration (and live migration) support, read the man page
1096 for ``gnt-instance migrate``
1097 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1098 has some small improvements in handling its state file
1104 *(Released Fri, 13 Jun 2008)*
1106 - Experimental readonly, REST-based remote API implementation;
1107 automatically started on master node, TCP port 5080, if enabled by
1108 ``--enable-rapi`` parameter to configure script.
1109 - Instance allocator support. Add and import instance accept a
1110 ``--iallocator`` parameter, and call that instance allocator to decide
1111 which node to use for the instance. The iallocator document describes
1112 what's expected from an allocator script.
1113 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1114 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1115 node is lost there is still enough memory to fail over the instances
1117 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1118 ``gnt-cluster verify``, to check for site-specific compliance. All the
1119 hooks will run, and their output, if any, will be displayed. Any
1120 failing hook will make the verification return an error value.
1121 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1122 primary and secondary interfaces
1123 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1124 that is still declared as part of the cluster and has failed.
1125 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1126 specifying output fields, that just adds the chosen fields to the
1128 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1129 export from the filesystem.
1130 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1131 have been added. Using them you can enable/disable acpi and pae
1132 support, and specify a path for a cd image to be exported to the
1133 instance. These parameters as the name suggest only work on HVM
1135 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1136 PAE support will be set to the previously hardcoded values, but the
1137 (previously hardcoded) path to the CDROM ISO image will be unset and
1138 if required, needs to be set manually with ``gnt-instance modify``
1140 - The address to which an instance's VNC console is bound is now
1141 selectable per-instance, rather than being cluster wide. Of course
1142 this only applies to instances controlled via VNC, so currently just
1143 applies to HVM clusters.
1149 *(Released Mon, 18 Feb 2008)*
1151 - more tweaks to the disk activation code (especially helpful for DRBD)
1152 - change the default ``gnt-instance list`` output format, now there is
1153 one combined status field (see the manpage for the exact values this
1155 - some more fixes for the mac export to hooks change
1156 - make Ganeti not break with DRBD 8.2.x (which changed the version
1157 format in ``/proc/drbd``) (issue 24)
1158 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1159 template, allowing migration from DRBD0.7+MD to DRBD8
1165 *(Released Wed, 30 Jan 2008)*
1167 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1169 - add command aliases infrastructure and a few aliases
1170 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1171 man pages and the ``--help`` option of ``gnt-node
1172 list``/``gnt-instance list``
1173 - fix ``gnt-backup list`` with down nodes (issue 21)
1174 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1175 - fix the dist archive and add a check for including svn/git files in
1177 - some developer-related changes: improve the burnin and the QA suite,
1178 add an upload script for testing during development
1184 *(Released Wed, 16 Jan 2008)*
1186 - experimental HVM support, read the install document, section
1187 "Initializing the cluster"
1188 - allow for the PVM hypervisor per-instance kernel and initrd paths
1189 - add a new command ``gnt-cluster verify-disks`` which uses a new
1190 algorithm to improve the reconnection of the DRBD pairs if the device
1191 on the secondary node has gone away
1192 - make logical volume code auto-activate LVs at disk activation time
1193 - slightly improve the speed of activating disks
1194 - allow specification of the MAC address at instance creation time, and
1195 changing it later via ``gnt-instance modify``
1196 - fix handling of external commands that generate lots of output on
1198 - update documentation with regard to minimum version of DRBD8 supported
1204 *(Released Tue, 4 Dec 2007)*
1206 - Log the ``xm create`` output to the node daemon log on failure (to
1207 help diagnosing the error)
1208 - In debug mode, log all external commands output if failed to the logs
1209 - Change parsing of lvm commands to ignore stderr
1215 *(Released Wed, 28 Nov 2007)*
1217 - Another round of updates to the DRBD 8 code to deal with more failures
1218 in the replace secondary node operation
1219 - Some more logging of failures in disk operations (lvm, drbd)
1220 - A few documentation updates
1227 *(Released Tue, 13 Nov 2007)*
1229 - Change configuration file format from Python's Pickle to JSON.
1230 Upgrading is possible using the cfgupgrade utility.
1231 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1232 faster replace disks and is more stable (DRBD 8 has many improvements
1233 compared to DRBD 0.7)
1234 - Added command line tags support (see man pages for ``gnt-instance``,
1235 ``gnt-node``, ``gnt-cluster``)
1236 - Added instance rename support
1237 - Added multi-instance startup/shutdown
1238 - Added cluster rename support
1239 - Added ``gnt-node evacuate`` to simplify some node operations
1240 - Added instance reboot operation that can speedup reboot as compared to
1242 - Soften the requirement that hostnames are in FQDN format
1243 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1245 - Removed dependency on debian's patched fping that uses the
1246 non-standard ``-S`` option
1247 - Now the OS definitions are searched for in multiple, configurable
1248 paths (easier for distros to package)
1249 - Some changes to the hooks infrastructure (especially the new
1250 post-configuration update hook)
1251 - Other small bugfixes
1253 .. vim: set textwidth=72 syntax=rst :