7 *(Released Tue, 2 Nov 2010)*
9 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
10 - Fixed master failover race with watcher
11 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
13 - Accept previously stopped instance for export with instance removal
14 - Simplify and extend the environment variables for instance OS scripts
15 - Added new node flags, ``master_capable`` and ``vm_capable``
16 - Added optional instance disk wiping prior during allocation. This is a
17 cluster-wide option and can be set/modified using
18 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
19 - Added IPv6 support, see :doc:`design document <design-2.3>` and
21 - Added a new watcher option (``--ignore-pause``)
22 - Added option to ignore offline node on instance start/stop
23 (``--ignore-offline``)
24 - Allow overriding OS parameters with ``gnt-instance reinstall``
25 - Added ability to change node's secondary IP address using ``gnt-node
27 - Implemented privilege separation for all daemons except
28 ``ganeti-noded``, see ``configure`` options
29 - Complain if an instance's disk is marked faulty in ``gnt-cluster
31 - Implemented job priorities (see ``ganeti(7)`` manpage)
32 - Ignore failures while shutting down instances during failover from
34 - Exit daemon's bootstrap process only once daemon is ready
35 - Export more information via ``LUQueryInstances``/remote API
36 - Improved documentation, QA and unittests
37 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
38 restart if the file was created or changed
39 - Added LUXI protocol version sent with each request and response,
40 allowing detection of server/client mismatches
41 - Moved the Python scripts among gnt-* and ganeti-* into modules
42 - Moved all code related to setting up SSH to an external script,
44 - Infrastructure changes for node group support in future versions
50 *(Released Fri, 19 Nov 2010)*
52 A few small bugs fixed, and some improvements to the build system:
54 - Fix documentation regarding conversion to drbd
55 - Fix validation of parameters in cluster modify (``gnt-cluster modify
57 - Fix error handling in node modify with multiple changes
58 - Allow remote imports without checked names
64 *(Released Tue, 19 Oct 2010)*
66 - Disable SSL session ID cache in RPC client
72 *(Released Thu, 14 Oct 2010)*
74 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
75 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
77 - Fix problems with interaction between Curl and Python's HTTP server,
78 resulting in increased speed in many RPC calls
79 - Improve our release script to prevent breakage with older aclocal and
86 *(Released Thu, 7 Oct 2010)*
88 - Fixed issue 125, replace hardcoded “xenvg” in ``gnt-cluster`` with
89 value retrieved from master
90 - Added support for blacklisted or hidden OS definitions
91 - Added simple lock monitor (accessible via (``gnt-debug locks``)
92 - Added support for -mem-path in KVM hypervisor abstraction layer
93 - Allow overriding instance parameters in tool for inter-cluster
94 instance moves (``tools/move-instance``)
95 - Improved opcode summaries (e.g. in ``gnt-job list``)
96 - Improve consistency of OS listing by sorting it
97 - Documentation updates
103 *(Released Fri, 8 Oct 2010)*
105 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
111 *(Released Mon, 4 Oct 2010)*
113 - Fixed regression in ``gnt-instance rename``
119 *(Released Wed, 22 Sep 2010)*
121 - Fixed OS_VARIANT variable for OS scripts
122 - Fixed cluster tag operations via RAPI
123 - Made ``setup-ssh`` exit with non-zero code if an error occurred
124 - Disabled RAPI CA checks in watcher
130 *(Released Mon, 23 Aug 2010)*
132 - Support DRBD versions of the format "a.b.c.d"
134 - Re-introduce support for usage from multiple threads in RAPI client
135 - Instance renames and modify via RAPI
136 - Work around race condition between processing and archival in job
138 - Mark opcodes following failed one as failed, too
139 - Job field ``lock_status`` was removed due to difficulties making it
140 work with the changed job queue in Ganeti 2.2; a better way to monitor
141 locks is expected for a later 2.2.x release
142 - Fixed dry-run behaviour with many commands
143 - Support ``ssh-agent`` again when adding nodes
144 - Many additional bugfixes
150 *(Released Fri, 30 Jul 2010)*
152 Important change: the internal RPC mechanism between Ganeti nodes has
153 changed from using a home-grown http library (based on the Python base
154 libraries) to use the PycURL library. This requires that PycURL is
155 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
156 against GnuTLS by default. cURL's support for GnuTLS had known issues
157 before cURL 7.21.0 and we recommend using the latest cURL release or
158 linking against OpenSSL. Most other distributions already link PycURL
159 and cURL against OpenSSL. The command::
161 python -c 'import pycurl; print pycurl.version'
163 can be used to determine the libraries PycURL and cURL are linked
166 Other significant changes:
168 - Rewrote much of the internals of the job queue, in order to achieve
169 better parallelism; this decouples job query operations from the job
170 processing, and it should allow much nicer behaviour of the master
171 daemon under load, and it also has uncovered some long-standing bugs
172 related to the job serialisation (now fixed)
173 - Added a default iallocator setting to the cluster parameters,
174 eliminating the need to always pass nodes or an iallocator for
175 operations that require selection of new node(s)
176 - Added experimental support for the LXC virtualization method
177 - Added support for OS parameters, which allows the installation of
178 instances to pass parameter to OS scripts in order to customise the
180 - Added a hypervisor parameter controlling the migration type (live or
181 non-live), since hypervisors have various levels of reliability; this
182 has renamed the 'live' parameter to 'mode'
183 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
184 logical volumes, meaning that cluster verify will ignore them and not
185 flag their presence as errors
186 - The watcher will now reset the error count for failed instances after
187 8 hours, thus allowing self-healing if the problem that caused the
188 instances to be down/fail to start has cleared in the meantime
189 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
190 check for, and warn, if the drbd module parameter ``usermode_helper``
191 is not consistent with the cluster-wide setting; this is needed to
192 make diagnose easier of failed drbd creations
193 - Started adding base IPv6 support, but this is not yet
194 enabled/available for use
195 - Rename operations (cluster, instance) will now return the new name,
196 which is especially useful if a short name was passed in
197 - Added support for instance migration in RAPI
198 - Added a tool to pre-configure nodes for the SSH setup, before joining
199 them to the cluster; this will allow in the future a simplified model
200 for node joining (but not yet fully enabled in 2.2); this needs the
201 paramiko python library
202 - Fixed handling of name-resolving errors
203 - Fixed consistency of job results on the error path
204 - Fixed master-failover race condition when executed multiple times in
206 - Fixed many bugs related to the job queue (mostly introduced during the
207 2.2 development cycle, so not all are impacting 2.1)
208 - Fixed instance migration with missing disk symlinks
209 - Fixed handling of unknown jobs in ``gnt-job archive``
210 - And many other small fixes/improvements
214 - Enhanced both the unittest and the QA coverage
215 - Switched the opcode validation to a generic model, and extended the
216 validation to all opcode parameters
217 - Changed more parts of the code that write shell scripts to use the
219 - Switched the master daemon to use the asyncore library for the Luxi
226 *(Released Thu, 17 Jun 2010)*
228 - Added tool (``move-instance``) and infrastructure to move instances
229 between separate clusters (see :doc:`separate documentation
230 <move-instance>` and :doc:`design document <design-2.2>`)
231 - Added per-request RPC timeout
232 - RAPI now requires a Content-Type header for requests with a body (e.g.
233 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
234 :rfc:`2616` (HTTP/1.1), section 7.2.1)
235 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
237 - Implemented initial support for running Ganeti daemons as separate
238 users, see configure-time flags ``--with-user-prefix`` and
239 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
241 - Instances can be removed after export (``gnt-backup export
243 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
244 key (instead of 1024 bit)
245 - Added new cluster configuration file for cluster domain secret
246 - Import/export now use SSL instead of SSH
247 - Added support for showing estimated time when exporting an instance,
248 see the ``ganeti-os-interface(7)`` manpage and look for
255 *(Released Tue, 16 Nov 2010)*
257 Some more bugfixes. Unless critical bugs occur, this will be the last
260 - Fix case of MAC special-values
261 - Fix mac checker regex
262 - backend: Fix typo causing “out of range” error
263 - Add missing --units in gnt-instance list man page
269 *(Released Tue, 24 Aug 2010)*
272 - Don't ignore secondary node silently on non-mirrored disk templates
274 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
275 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
276 - Properly document the usb_mouse parameter
277 - Fix path in ganeti-rapi(8) (issue 116)
278 - Adjust error message when the ganeti user's .ssh directory is
280 - Add same-node-check when changing the disk template to drbd
286 *(Released Fri, 16 Jul 2010)*
289 - Add an option to only select some reboot types during qa/burnin.
290 (on some hypervisors consequent reboots are not supported)
291 - Fix infrequent race condition in master failover. Sometimes the old
292 master ip address would be still detected as up for a short time
293 after it was removed, causing failover to fail.
294 - Decrease mlockall warnings when the ctypes module is missing. On
295 Python 2.4 we support running even if no ctypes module is installed,
296 but we were too verbose about this issue.
297 - Fix building on old distributions, on which man doesn't have a
299 - Fix RAPI not to ignore the MAC address on instance creation
300 - Implement the old instance creation format in the RAPI client.
306 *(Released Thu, 01 Jul 2010)*
308 A small bugfix release:
309 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
310 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
311 a non-existing option
312 - Fix instance create: the --force-variant option was ignored
313 - Improve pylint 0.21 compatibility and warnings with Python 2.6
314 - Fix modify node storage with non-FQDN arguments
315 - Fix RAPI client to authenticate under Python 2.6 when used
316 for more than 5 requests needing authentication
317 - Fix gnt-instance modify -t (storage) giving a wrong error message
318 when converting a non-shutdown drbd instance to plain
324 *(Released Fri, 18 Jun 2010)*
326 A small bugfix release:
328 - Fix live migration of KVM instances started with older Ganeti
329 versions which had fewer hypervisor parameters
330 - Fix gnt-instance grow-disk on down instances
331 - Fix an error-reporting bug during instance migration
332 - Better checking of the ``--net`` and ``--disk`` values, to avoid
333 silently ignoring broken ones
334 - Fix an RPC error reporting bug affecting, for example, RAPI client
336 - Fix bug triggered by different API version os-es on different nodes
337 - Fix a bug in instance startup with custom hvparams: OS level
338 parameters would fail to be applied.
339 - Fix the RAPI client under Python 2.6 (but more work is needed to
340 make it work completely well with OpenSSL)
341 - Fix handling of errors when resolving names from DNS
347 *(Released Thu, 3 Jun 2010)*
349 A medium sized development cycle. Some new features, and some
350 fixes/small improvements/cleanups.
355 The node deamon now tries to mlock itself into memory, unless the
356 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
357 its logs, and falls back to console logging. This allows emergency
358 features such as ``gnt-node powercycle`` to work even in the event of a
359 broken node disk (tested offlining the disk hosting the node's
360 filesystem and dropping its memory caches; don't try this at home)
362 KVM: add vhost-net acceleration support. It can be tested with a new
363 enough version of the kernel and of qemu-kvm.
365 KVM: Add instance chrooting feature. If you use privilege dropping for
366 your VMs you can also now force them to chroot to an empty directory,
367 before starting the emulated guest.
369 KVM: Add maximum migration bandwith and maximum downtime tweaking
370 support (requires a new-enough version of qemu-kvm).
372 Cluster verify will now warn if the master node doesn't have the master
375 Add a new (incompatible) instance creation request format to RAPI which
376 supports all parameters (previously only a subset was supported, and it
377 wasn't possible to extend the old format to accomodate all the new
378 features. The old format is still supported, and a client can check for
379 this feature, before using it, by checking for its presence in the
380 ``features`` RAPI resource.
382 Now with ancient latin support. Try it passing the ``--roman`` option to
383 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
384 (requires the python-roman module to be installed, in order to work).
389 As usual many internal code refactorings, documentation updates, and
392 - Lots of improvements and cleanups to the experimental Remote API
393 (RAPI) client library.
394 - A new unit test suite for the core daemon libraries.
395 - A fix to creating missing directories makes sure the umask is not
396 applied anymore. This enforces the same directory permissions
398 - Better handling terminating daemons with ctrl+c (used when running
399 them in debugging mode).
400 - Fix a race condition in live migrating a KVM instance, when stat()
401 on the old proc status file returned EINVAL, which is an unexpected
403 - Fixed manpage checking with newer man and utf-8 charachters. But now
404 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
410 *(Released Fri, 7 May 2010)*
412 Fix a bug which prevented untagged KVM instances from starting.
418 *(Released Fri, 7 May 2010)*
420 Another release with a long development cycle, during which many
421 different features were added.
426 The KVM hypervisor now can run the individual instances as non-root, to
427 reduce the impact of a VM being hijacked due to bugs in the
428 hypervisor. It is possible to run all instances as a single (non-root)
429 user, to manually specify a user for each instance, or to dynamically
430 allocate a user out of a cluster-wide pool to each instance, with the
431 guarantee that no two instances will run under the same user ID on any
434 An experimental RAPI client library, that can be used standalone
435 (without the other Ganeti libraries), is provided in the source tree as
436 ``lib/rapi/client.py``. Note this client might change its interface in
437 the future, as we iterate on its capabilities.
439 A new command, ``gnt-cluster renew-crypto`` has been added to easily
440 replace the cluster's certificates and crypto keys. This might help in
441 case they have been compromised, or have simply expired.
443 A new disk option for instance creation has been added that allows one
444 to "adopt" currently existing logical volumes, with data
445 preservation. This should allow easier migration to Ganeti from
446 unmanaged (or managed via other software) instances.
448 Another disk improvement is the possibility to convert between redundant
449 (DRBD) and plain (LVM) disk configuration for an instance. This should
450 allow better scalability (starting with one node and growing the
451 cluster, or shrinking a two-node cluster to one node).
453 A new feature that could help with automated node failovers has been
454 implemented: if a node sees itself as offline (by querying the master
455 candidates), it will try to shutdown (hard) all instances and any active
456 DRBD devices. This reduces the risk of duplicate instances if an
457 external script automatically failovers the instances on such nodes. To
458 enable this, the cluster parameter ``maintain_node_health`` should be
459 enabled; in the future this option (per the name) will enable other
460 automatic maintenance features.
462 Instance export/import now will reuse the original instance
463 specifications for all parameters; that means exporting an instance,
464 deleting it and the importing it back should give an almost identical
465 instance. Note that the default import behaviour has changed from
466 before, where it created only one NIC; now it recreates the original
469 Cluster verify has added a few new checks: SSL certificates validity,
470 /etc/hosts consistency across the cluster, etc.
475 As usual, many internal changes were done, documentation fixes,
478 - Fixed cluster initialization with disabled cluster storage (regression
480 - File-based storage supports growing the disks
481 - Fixed behaviour of node role changes
482 - Fixed cluster verify for some corner cases, plus a general rewrite of
483 cluster verify to allow future extension with more checks
484 - Fixed log spamming by watcher and node daemon (regression introduced
486 - Fixed possible validation issues when changing the list of enabled
488 - Fixed cleanup of /etc/hosts during node removal
489 - Fixed RAPI response for invalid methods
490 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
491 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
492 from ide disks, etc.)
493 - Allow OS changes without re-installation (to record a changed OS
494 outside of Ganeti, or to allow OS renames)
495 - Allow instance creation without OS installation (useful for example if
496 the OS will be installed manually, or restored from a backup not in
498 - Implemented option to make cluster ``copyfile`` use the replication
500 - Added list of enabled hypervisors to ssconf (possibly useful for
502 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
504 - A partial form of node re-IP is possible via node readd, which now
505 allows changed node primary IP
506 - Command line utilities now show an informational message if the job is
508 - The logs of the master daemon now show the PID/UID/GID of the
515 *(Released Fri, 12 Mar 2010)*
517 During the 2.1.0 long release candidate cycle, a lot of improvements and
518 changes have accumulated with were released later as 2.1.1.
523 The node evacuate command (``gnt-node evacuate``) was significantly
524 rewritten, and as such the IAllocator protocol was changed - a new
525 request type has been added. This unfortunate change during a stable
526 series is designed to improve performance of node evacuations; on
527 clusters with more than about five nodes and which are well-balanced,
528 evacuation should proceed in parallel for all instances of the node
529 being evacuated. As such, any existing IAllocator scripts need to be
530 updated, otherwise the above command will fail due to the unknown
531 request. The provided "dumb" allocator has not been updated; but the
532 ganeti-htools package supports the new protocol since version 0.2.4.
534 Another important change is increased validation of node and instance
535 names. This might create problems in special cases, if invalid host
536 names are being used.
538 Also, a new layer of hypervisor parameters has been added, that sits at
539 OS level between the cluster defaults and the instance ones. This allows
540 customisation of virtualization parameters depending on the installed
541 OS. For example instances with OS 'X' may have a different KVM kernel
542 (or any other parameter) than the cluster defaults. This is intended to
543 help managing a multiple OSes on the same cluster, without manual
544 modification of each instance's parameters.
546 A tool for merging clusters, ``cluster-merge``, has been added in the
552 - Improved the int/float conversions that should make the code more
553 robust in face of errors from the node daemons
554 - Fixed the remove node code in case of internal configuration errors
555 - Fixed the node daemon behaviour in face of inconsistent queue
556 directory (e.g. read-only file-system where we can't open the files
558 - Fixed the behaviour of gnt-node modify for master candidate demotion;
559 now it either aborts cleanly or, if given the new “auto_promote”
560 parameter, will automatically promote other nodes as needed
561 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
562 completely prevent Ganeti from working
563 - Fixed bug for instance export when not all disks were successfully
565 - Fixed behaviour of node add when the new node is slow in starting up
567 - Fixed handling of signals in the LUXI client, which should improve
568 behaviour of command-line scripts
569 - Added checks for invalid node/instance names in the configuration (now
570 flagged during cluster verify)
571 - Fixed watcher behaviour for disk activation errors
572 - Fixed two potentially endless loops in http library, which led to the
573 RAPI daemon hanging and consuming 100% CPU in some cases
574 - Fixed bug in RAPI daemon related to hashed passwords
575 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
577 - Enhanced compatibility with non-Debian OSes, but not using absolute
578 path in some commands and allowing customisation of the ssh
579 configuration directory
580 - Fixed possible future issue with new Python versions by abiding to the
581 proper use of ``__slots__`` attribute on classes
582 - Added checks that should prevent directory traversal attacks
583 - Many documentation fixes based on feedback from users
588 - Added an “early_release” more for instance replace disks and node
589 evacuate, where we release locks earlier and thus allow higher
590 parallelism within the cluster
591 - Added watcher hooks, intended to allow the watcher to restart other
592 daemons (e.g. from the ganeti-nbma project), but they can be used of
593 course for any other purpose
594 - Added a compile-time disable for DRBD barriers, to increase
595 performance if the administrator trusts the power supply or the
596 storage system to not lose writes
597 - Added the option of using syslog for logging instead of, or in
598 addition to, Ganeti's own log files
599 - Removed boot restriction for paravirtual NICs for KVM, recent versions
600 can indeed boot from a paravirtual NIC
601 - Added a generic debug level for many operations; while this is not
602 used widely yet, it allows one to pass the debug value all the way to
604 - Enhanced the hooks environment for instance moves (failovers,
605 migrations) where the primary/secondary nodes changed during the
606 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
607 - Enhanced data validations for many user-supplied values; one important
608 item is the restrictions imposed on instance and node names, which
609 might reject some (invalid) host names
610 - Add a configure-time option to disable file-based storage, if it's not
611 needed; this allows greater security separation between the master
612 node and the other nodes from the point of view of the inter-node RPC
614 - Added user notification in interactive tools if job is waiting in the
615 job queue or trying to acquire locks
616 - Added log messages when a job is waiting for locks
617 - Added filtering by node tags in instance operations which admit
618 multiple instances (start, stop, reboot, reinstall)
619 - Added a new tool for cluster mergers, ``cluster-merge``
620 - Parameters from command line which are of the form ``a=b,c=d`` can now
621 use backslash escapes to pass in values which contain commas,
622 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
624 - For KVM, the instance name is the first parameter passed to KVM, so
625 that it's more visible in the process list
631 *(Released Tue, 2 Mar 2010)*
633 Ganeti 2.1 brings many improvements with it. Major changes:
635 - Added infrastructure to ease automated disk repairs
636 - Added new daemon to export configuration data in a cheaper way than
638 - Instance NICs can now be routed instead of being associated with a
640 - Improved job locking logic to reduce impact of jobs acquiring multiple
641 locks waiting for other long-running jobs
643 In-depth implementation details can be found in the Ganeti 2.1 design
649 - Added chroot hypervisor
650 - Added more options to xen-hvm hypervisor (``kernel_path`` and
652 - Added more options to xen-pvm hypervisor (``use_bootloader``,
653 ``bootloader_path`` and ``bootloader_args``)
654 - Added the ``use_localtime`` option for the xen-hvm and kvm
655 hypervisors, and the default value for this has changed to false (in
656 2.0 xen-hvm always enabled it)
657 - Added luxi call to submit multiple jobs in one go
658 - Added cluster initialization option to not modify ``/etc/hosts``
660 - Added network interface parameters
661 - Added dry run mode to some LUs
662 - Added RAPI resources:
664 - ``/2/instances/[instance_name]/info``
665 - ``/2/instances/[instance_name]/replace-disks``
666 - ``/2/nodes/[node_name]/evacuate``
667 - ``/2/nodes/[node_name]/migrate``
668 - ``/2/nodes/[node_name]/role``
669 - ``/2/nodes/[node_name]/storage``
670 - ``/2/nodes/[node_name]/storage/modify``
671 - ``/2/nodes/[node_name]/storage/repair``
673 - Added OpCodes to evacuate or migrate all instances on a node
674 - Added new command to list storage elements on nodes (``gnt-node
675 list-storage``) and modify them (``gnt-node modify-storage``)
676 - Added new ssconf files with master candidate IP address
677 (``ssconf_master_candidates_ips``), node primary IP address
678 (``ssconf_node_primary_ips``) and node secondary IP address
679 (``ssconf_node_secondary_ips``)
680 - Added ``ganeti-confd`` and a client library to query the Ganeti
681 configuration via UDP
682 - Added ability to run hooks after cluster initialization and before
684 - Added automatic mode for disk replace (``gnt-instance replace-disks
686 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
687 after catastrophic data-loss
688 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
690 - Added ``gnt-instance move`` command to move instances
691 - Added ``gnt-cluster watcher`` command to control watcher
692 - Added ``gnt-node powercycle`` command to powercycle nodes
693 - Added new job status field ``lock_status``
694 - Added parseable error codes to cluster verification (``gnt-cluster
695 verify --error-codes``) and made output less verbose (use
696 ``--verbose`` to restore previous behaviour)
697 - Added UUIDs to the main config entities (cluster, nodes, instances)
698 - Added support for OS variants
699 - Added support for hashed passwords in the Ganeti remote API users file
701 - Added option to specify maximum timeout on instance shutdown
702 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
703 - Added new helper script to start and stop Ganeti daemons
704 (``daemon-util``), with the intent to reduce the work necessary to
705 adjust Ganeti for non-Debian distributions and to start/stop daemons
707 - Added more unittests
708 - Fixed critical bug in ganeti-masterd startup
709 - Removed the configure-time ``kvm-migration-port`` parameter, this is
710 now customisable at the cluster level for both the KVM and Xen
711 hypervisors using the new ``migration_port`` parameter
712 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
713 reinstalling an instance
714 - Allowed ``@`` in tag names
715 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
716 - Many documentation updates
717 - Distribute hypervisor files on ``gnt-cluster redist-conf``
718 - ``gnt-instance reinstall`` can now reinstall multiple instances
719 - Updated many command line parameters
720 - Introduced new OS API version 15
721 - No longer support a default hypervisor
722 - Treat virtual LVs as inexistent
723 - Improved job locking logic to reduce lock contention
724 - Match instance and node names case insensitively
725 - Reimplemented bash completion script to be more complete
732 *(Released Thu, 4 Feb 2010)*
734 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
735 - Fix a string formatting bug
736 - Improve safety of the code in some error paths
737 - Improve data validation in the master of values returned from nodes
743 *(Released Thu, 17 Dec 2009)*
745 - Fix security issue due to missing validation of iallocator names; this
746 allows local and remote execution of arbitrary executables
747 - Fix failure of gnt-node list during instance removal
748 - Ship the RAPI documentation in the archive
754 *(Released Wed, 30 Sep 2009)*
756 - Fixed many wrong messages
757 - Fixed a few bugs related to the locking library
758 - Fixed MAC checking at instance creation time
759 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
760 - Fixed a few issues related to signal handling in both daemons and
762 - Fixed the example startup script provided
763 - Fixed insserv dependencies in the example startup script (patch from
765 - Fixed handling of drained nodes in the iallocator framework
766 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
768 - Fixed error related to invalid job IDs in job polling
769 - Fixed job/opcode persistence on unclean master shutdown
770 - Fixed handling of partial job processing after unclean master
772 - Fixed error reporting from LUs, previously all errors were converted
773 into execution errors
774 - Fixed error reporting from burnin
775 - Decreased significantly the memory usage of the job queue
776 - Optimised slightly multi-job submission
777 - Optimised slightly opcode loading
778 - Backported the multi-job submit framework from the development
779 branch; multi-instance start and stop should be faster
780 - Added script to clean archived jobs after 21 days; this will reduce
781 the size of the queue directory
782 - Added some extra checks in disk size tracking
783 - Added an example ethers hook script
784 - Added a cluster parameter that prevents Ganeti from modifying of
786 - Added more node information to RAPI responses
787 - Added a ``gnt-job watch`` command that allows following the ouput of a
789 - Added a bind-address option to ganeti-rapi
790 - Added more checks to the configuration verify
791 - Enhanced the burnin script such that some operations can be retried
793 - Converted instance reinstall to multi-instance model
799 *(Released Fri, 7 Aug 2009)*
801 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
802 to allow using the pre-2.0.2 behaviour in activation, if any existing
803 instances have mismatched disk sizes in the configuration
804 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
805 any configuration mismatches for disk sizes
806 - Added ``gnt-master cluste-failover --no-voting`` to allow master
807 failover to work on two-node clusters
808 - Fixed the ``--net`` option of ``gnt-backup import``, which was
810 - Fixed detection of OS script errors in ``gnt-backup export``
811 - Fixed exit code of ``gnt-backup export``
817 *(Released Fri, 17 Jul 2009)*
819 - Added experimental support for stripped logical volumes; this should
820 enhance performance but comes with a higher complexity in the block
821 device handling; stripping is only enabled when passing
822 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
823 affected even in the non-stripped mode
824 - Improved resiliency against transient failures at the end of DRBD
825 resyncs, and in general of DRBD resync checks
826 - Fixed a couple of issues with exports and snapshot errors
827 - Fixed a couple of issues in instance listing
828 - Added display of the disk size in ``gnt-instance info``
829 - Fixed checking for valid OSes in instance creation
830 - Fixed handling of the "vcpus" parameter in instance listing and in
831 general of invalid parameters
832 - Fixed http server library, and thus RAPI, to handle invalid
833 username/password combinations correctly; this means that now they
834 report unauthorized for queries too, not only for modifications,
835 allowing earlier detect of configuration problems
836 - Added a new "role" node list field, equivalent to the master/master
837 candidate/drained/offline flags combinations
838 - Fixed cluster modify and changes of candidate pool size
839 - Fixed cluster verify error messages for wrong files on regular nodes
840 - Fixed a couple of issues with node demotion from master candidate role
841 - Fixed node readd issues
842 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
843 - Added a new ``--no-voting`` option for masterfailover to fix failover
844 on two-nodes clusters when the former master node is unreachable
845 - Added instance reinstall over RAPI
851 *(Released Tue, 16 Jun 2009)*
853 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
854 allow re-adding the start in single-user option (regression from 1.2)
855 - the watcher writes the instance status to a file, to allow monitoring
856 to report the instance status (from the master) based on cached
857 results of the watcher's queries; while this can get stale if the
858 watcher is being locked due to other work on the cluster, this is
860 - the watcher now also restarts the node daemon and the rapi daemon if
862 - fixed the watcher to handle full and drained queue cases
863 - hooks export more instance data in the environment, which helps if
864 hook scripts need to take action based on the instance's properties
865 (no longer need to query back into ganeti)
866 - instance failovers when the instance is stopped do not check for free
867 RAM, so that failing over a stopped instance is possible in low memory
869 - rapi uses queries for tags instead of jobs (for less job traffic), and
870 for cluster tags it won't talk to masterd at all but read them from
872 - a couple of error handling fixes in RAPI
873 - drbd handling: improved the error handling of inconsistent disks after
874 resync to reduce the frequency of "there are some degraded disks for
875 this instance" messages
876 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
877 error handling path called a wrong function name)
883 *(Released Wed, 27 May 2009)*
885 - no changes from rc5
888 Version 2.0 release candidate 5
889 -------------------------------
891 *(Released Wed, 20 May 2009)*
893 - fix a couple of bugs (validation, argument checks)
894 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
895 - some small improvements to RAPI and IAllocator
896 - make watcher automatically start the master daemon if down
899 Version 2.0 release candidate 4
900 -------------------------------
902 *(Released Mon, 27 Apr 2009)*
904 - change the OS list to not require locks; this helps with big clusters
905 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
906 volume group is broken
907 - ``gnt-instance info``, without any arguments, doesn't run for all
908 instances anymore; either pass ``--all`` or pass the desired
909 instances; this helps against mistakes on big clusters where listing
910 the information for all instances takes a long time
911 - miscellaneous doc and man pages fixes
914 Version 2.0 release candidate 3
915 -------------------------------
917 *(Released Wed, 8 Apr 2009)*
919 - Change the internal locking model of some ``gnt-node`` commands, in
920 order to reduce contention (and blocking of master daemon) when
921 batching many creation/reinstall jobs
922 - Fixes to Xen soft reboot
923 - No longer build documentation at build time, instead distribute it in
924 the archive, in order to reduce the need for the whole docbook/rst
928 Version 2.0 release candidate 2
929 -------------------------------
931 *(Released Fri, 27 Mar 2009)*
933 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
934 - Fix watcher startup sequence, improves the behaviour of busy clusters
935 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
936 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
937 bind address and other places
938 - Some documentation fixes and updates
941 Version 2.0 release candidate 1
942 -------------------------------
944 *(Released Mon, 2 Mar 2009)*
946 - More documentation updates, now all docs should be more-or-less
948 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
950 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
952 - Fix an issue related to $libdir/run/ganeti and cluster creation
958 *(Released Thu, 19 Feb 2009)*
960 - Xen PVM and KVM have switched the default value for the instance root
961 disk to the first partition on the first drive, instead of the whole
962 drive; this means that the OS installation scripts must be changed
964 - Man pages have been updated
965 - RAPI has been switched by default to HTTPS, and the exported functions
966 should all work correctly
967 - RAPI v1 has been removed
968 - Many improvements to the KVM hypervisor
969 - Block device errors are now better reported
970 - Many other bugfixes and small improvements
976 *(Released Mon, 26 Jan 2009)*
978 - Version 2 is a general rewrite of the code and therefore the
979 differences are too many to list, see the design document for 2.0 in
980 the ``doc/`` subdirectory for more details
981 - In this beta version there is not yet a migration path from 1.2 (there
982 will be one in the final 2.0 release)
983 - A few significant changes are:
985 - all commands are executed by a daemon (``ganeti-masterd``) and the
986 various ``gnt-*`` commands are just front-ends to it
987 - all the commands are entered into, and executed from a job queue,
988 see the ``gnt-job(8)`` manpage
989 - the RAPI daemon supports read-write operations, secured by basic
990 HTTP authentication on top of HTTPS
991 - DRBD version 0.7 support has been removed, DRBD 8 is the only
992 supported version (when migrating from Ganeti 1.2 to 2.0, you need
993 to migrate to DRBD 8 first while still running Ganeti 1.2)
994 - DRBD devices are using statically allocated minor numbers, which
995 will be assigned to existing instances during the migration process
996 - there is support for both Xen PVM and Xen HVM instances running on
998 - KVM virtualization is supported too
999 - file-based storage has been implemented, which means that it is
1000 possible to run the cluster without LVM and DRBD storage, for
1001 example using a shared filesystem exported from shared storage (and
1002 still have live migration)
1008 *(Released Tue, 13 Jan 2009)*
1010 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1011 - Reuse the old instance mac address by default on instance import, if
1012 the instance name is the same.
1013 - Handle situations in which the node info rpc returns incomplete
1015 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1016 - Improved version of batcher:
1018 - state file support
1019 - instance mac address support
1020 - support for HVM clusters/instances
1022 - Add an option to show the number of cpu sockets and nodes in
1024 - Support OSes that handle more than one version of the OS api (but do
1025 not change the current API in any other way)
1026 - Fix ``gnt-node migrate``
1027 - ``gnt-debug`` man page
1028 - Fixes various more typos and small issues
1029 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1035 *(Released Wed, 24 Sep 2008)*
1037 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1038 type of disk exported to fully virtualized instances.
1039 - provide access to the serial console of HVM instances
1040 - instance auto_balance flag, set by default. If turned off it will
1041 avoid warnings on cluster verify if there is not enough memory to fail
1042 over an instance. in the future it will prevent automatically failing
1043 it over when we will support that.
1044 - batcher tool for instance creation, see ``tools/README.batcher``
1045 - ``gnt-instance reinstall --select-os`` to interactively select a new
1046 operating system when reinstalling an instance.
1047 - when changing the memory amount on instance modify a check has been
1048 added that the instance will be able to start. also warnings are
1049 emitted if the instance will not be able to fail over, if auto_balance
1051 - documentation fixes
1052 - sync fields between ``gnt-instance list/modify/add/import``
1053 - fix a race condition in drbd when the sync speed was set after giving
1054 the device a remote peer.
1060 *(Released Tue, 22 Jul 2008)*
1062 - note: the allowed size and number of tags per object were reduced
1063 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1064 - fixed twisted 8.x compatibility
1065 - fixed ``gnt-instance replace-disks`` with iallocator
1066 - add TCP keepalives on twisted connections to detect restarted nodes
1067 - disk increase support, see ``gnt-instance grow-disk``
1068 - implement bulk node/instance query for RAPI
1069 - add tags in node/instance listing (optional)
1070 - experimental migration (and live migration) support, read the man page
1071 for ``gnt-instance migrate``
1072 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1073 has some small improvements in handling its state file
1079 *(Released Fri, 13 Jun 2008)*
1081 - Experimental readonly, REST-based remote API implementation;
1082 automatically started on master node, TCP port 5080, if enabled by
1083 ``--enable-rapi`` parameter to configure script.
1084 - Instance allocator support. Add and import instance accept a
1085 ``--iallocator`` parameter, and call that instance allocator to decide
1086 which node to use for the instance. The iallocator document describes
1087 what's expected from an allocator script.
1088 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1089 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1090 node is lost there is still enough memory to fail over the instances
1092 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1093 ``gnt-cluster verify``, to check for site-specific compliance. All the
1094 hooks will run, and their output, if any, will be displayed. Any
1095 failing hook will make the verification return an error value.
1096 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1097 primary and secondary interfaces
1098 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1099 that is still declared as part of the cluster and has failed.
1100 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1101 specifying output fields, that just adds the chosen fields to the
1103 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1104 export from the filesystem.
1105 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1106 have been added. Using them you can enable/disable acpi and pae
1107 support, and specify a path for a cd image to be exported to the
1108 instance. These parameters as the name suggest only work on HVM
1110 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1111 PAE support will be set to the previously hardcoded values, but the
1112 (previously hardcoded) path to the CDROM ISO image will be unset and
1113 if required, needs to be set manually with ``gnt-instance modify``
1115 - The address to which an instance's VNC console is bound is now
1116 selectable per-instance, rather than being cluster wide. Of course
1117 this only applies to instances controlled via VNC, so currently just
1118 applies to HVM clusters.
1124 *(Released Mon, 18 Feb 2008)*
1126 - more tweaks to the disk activation code (especially helpful for DRBD)
1127 - change the default ``gnt-instance list`` output format, now there is
1128 one combined status field (see the manpage for the exact values this
1130 - some more fixes for the mac export to hooks change
1131 - make Ganeti not break with DRBD 8.2.x (which changed the version
1132 format in ``/proc/drbd``) (issue 24)
1133 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1134 template, allowing migration from DRBD0.7+MD to DRBD8
1140 *(Released Wed, 30 Jan 2008)*
1142 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1144 - add command aliases infrastructure and a few aliases
1145 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1146 man pages and the ``--help`` option of ``gnt-node
1147 list``/``gnt-instance list``
1148 - fix ``gnt-backup list`` with down nodes (issue 21)
1149 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1150 - fix the dist archive and add a check for including svn/git files in
1152 - some developer-related changes: improve the burnin and the QA suite,
1153 add an upload script for testing during development
1159 *(Released Wed, 16 Jan 2008)*
1161 - experimental HVM support, read the install document, section
1162 "Initializing the cluster"
1163 - allow for the PVM hypervisor per-instance kernel and initrd paths
1164 - add a new command ``gnt-cluster verify-disks`` which uses a new
1165 algorithm to improve the reconnection of the DRBD pairs if the device
1166 on the secondary node has gone away
1167 - make logical volume code auto-activate LVs at disk activation time
1168 - slightly improve the speed of activating disks
1169 - allow specification of the MAC address at instance creation time, and
1170 changing it later via ``gnt-instance modify``
1171 - fix handling of external commands that generate lots of output on
1173 - update documentation with regard to minimum version of DRBD8 supported
1179 *(Released Tue, 4 Dec 2007)*
1181 - Log the ``xm create`` output to the node daemon log on failure (to
1182 help diagnosing the error)
1183 - In debug mode, log all external commands output if failed to the logs
1184 - Change parsing of lvm commands to ignore stderr
1190 *(Released Wed, 28 Nov 2007)*
1192 - Another round of updates to the DRBD 8 code to deal with more failures
1193 in the replace secondary node operation
1194 - Some more logging of failures in disk operations (lvm, drbd)
1195 - A few documentation updates
1202 *(Released Tue, 13 Nov 2007)*
1204 - Change configuration file format from Python's Pickle to JSON.
1205 Upgrading is possible using the cfgupgrade utility.
1206 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1207 faster replace disks and is more stable (DRBD 8 has many improvements
1208 compared to DRBD 0.7)
1209 - Added command line tags support (see man pages for ``gnt-instance``,
1210 ``gnt-node``, ``gnt-cluster``)
1211 - Added instance rename support
1212 - Added multi-instance startup/shutdown
1213 - Added cluster rename support
1214 - Added ``gnt-node evacuate`` to simplify some node operations
1215 - Added instance reboot operation that can speedup reboot as compared to
1217 - Soften the requirement that hostnames are in FQDN format
1218 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1220 - Removed dependency on debian's patched fping that uses the
1221 non-standard ``-S`` option
1222 - Now the OS definitions are searched for in multiple, configurable
1223 paths (easier for distros to package)
1224 - Some changes to the hooks infrastructure (especially the new
1225 post-configuration update hook)
1226 - Other small bugfixes
1228 .. vim: set textwidth=72 syntax=rst :
projects / ganeti-local / blob