9 - Added a new parameter to instance stop/start called ``--no-remember``
10 that will make the state change to not be remembered
16 *(Released Thu, 12 May 2011)*
18 Many bug-fixes and a few new small features:
20 - Fixed a bug related to log opening failures
21 - Fixed a bug in instance listing with orphan instances
22 - Fixed a bug which prevented resetting the cluster-level node parameter
23 ``oob_program`` to the default
24 - Many fixes related to the ``cluster-merge`` tool
25 - Fixed a race condition in the lock monitor, which caused failures
26 during (at least) creation of many instances in parallel
27 - Improved output for gnt-job info
28 - Removed the quiet flag on some ssh calls which prevented debugging
30 - Improved the N+1 failure messages in cluster verify by actually
31 showing the memory values (needed and available)
32 - Increased lock attempt timeouts so that when executing long operations
33 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
34 too early and thus prevent the use of the 'fair' mechanism
35 - Changed instance query data (``gnt-instance info``) to not acquire
36 locks unless needed, thus allowing its use on locked instance if only
37 static information is asked for
38 - Improved behaviour with filesystems that do not support rename on an
40 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
41 kept locks on all nodes during the wipe, which is unneeded
42 - Fixed ``gnt-watcher`` handling of errors during hooks execution
43 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
44 10GiB) which caused the wipe to fail right at the end in some cases
45 - Fixed master IP activation when doing master failover with no-voting
46 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
47 the master node itself
48 - Fixed potential data-loss in under disk full conditions, where Ganeti
49 wouldn't check correctly the return code and would consider
50 partially-written files 'correct'
51 - Fixed bug related to multiple VGs and DRBD disk replacing
52 - Added new disk parameter ``metavg`` that allows placement of the meta
53 device for DRBD in a different volume group
54 - Fixed error handling in the node daemon when the system libc doesn't
55 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
56 - Fixed lock release during replace-disks, which kept cluster-wide locks
57 when doing disk replaces with an iallocator script
58 - Added check for missing bridges in cluster verify
59 - Handle EPIPE errors while writing to the terminal better, so that
60 piping the output to e.g. ``less`` doesn't cause a backtrace
61 - Fixed rare case where a ^C during Luxi calls could have been
62 interpreted as server errors, instead of simply terminating
63 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
65 - Added a few more parameters to the KVM hypervisor, allowing a second
66 CDROM, custom disk type for CDROMs and a floppy image
67 - Removed redundant message in instance rename when the name is given
69 - Added option to ``gnt-instance recreate-disks`` to allow creating the
70 disks on new nodes, allowing recreation when the original instance
71 nodes are completely gone
72 - Added option when converting disk templates to DRBD to skip waiting
73 for the resync, in order to make the instance available sooner
74 - Added two new variables to the OS scripts environment (containing the
76 - Made the root_path and optional parameter for the xen-pvm hypervisor,
77 to allow use of ``pvgrub`` as bootloader
78 - Changed the instance memory modifications to only check out-of-memory
79 conditions on memory increases, and turned the secondary node warnings
80 into errors (they can still be overridden via ``--force``)
81 - Fixed the handling of a corner case when the Python installation gets
82 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
83 to execute a command that doesn't exist
84 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
85 node of the instance returned failures during instance shutdown; this
86 adds the option ``--ignore-consistency`` to gnt-instance move
88 And as usual, various improvements to the error messages, documentation
94 *(Released Wed, 09 Mar 2011)*
96 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
97 the RAPI users file if run twice (even with ``--dry-run``).
99 The release fixes that bug (nothing else changed).
105 *(Released Mon, 07 Mar 2011)*
107 Final 2.4.0 release. Just a few small fixes:
109 - Fixed RAPI node evacuate
110 - Fixed the kvm-ifup script
111 - Fixed internal error handling for special job cases
112 - Updated man page to specify the escaping feature for options
118 *(Released Mon, 28 Feb 2011)*
120 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
121 that this feature wiped the disks of the wrong instance, leading to loss
126 - Fixed title of query field containing instance name
127 - Expanded the glossary in the documentation
128 - Fixed one unittest (internal issue)
134 *(Released Mon, 21 Feb 2011)*
136 A number of bug fixes plus just a couple functionality changes.
138 On the user-visible side, the ``gnt-* list`` command output has changed
139 with respect to "special" field states. The current rc1 style of display
140 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
141 the default output mode special fields states are displayed as follows:
143 - Offline resource: ``*``
144 - Unavailable/not applicable: ``-``
145 - Data missing (RPC failure): ``?``
146 - Unknown field: ``??``
148 Another user-visible change is the addition of ``--force-join`` to
153 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
154 - Fixed regression in RAPI/instance reinstall where all parameters were
155 required (instead of optional)
156 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
157 - Fixed iallocator usage (offline nodes were not considered offline)
158 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
159 - Fixed hypervisor and OS parameter validation with respect to
161 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
163 - Fixed ``tools/listrunner`` with respect to agent-based usage
169 *(Released Fri, 4 Feb 2011)*
171 Many changes and fixes since the beta1 release. While there were some
172 internal changes, the code has been mostly stabilised for the RC
175 Note: the dumb allocator was removed in this release, as it was not kept
176 up-to-date with the IAllocator protocol changes. It is recommended to
177 use the ``hail`` command from the ganeti-htools package.
179 Note: the 2.4 and up versions of Ganeti are not compatible with the
180 0.2.x branch of ganeti-htools. You need to upgrade to
181 ganeti-htools-0.3.0 (or later).
183 Regressions fixed from 2.3
184 ~~~~~~~~~~~~~~~~~~~~~~~~~~
186 - Fixed the ``gnt-cluster verify-disks`` command
187 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
189 - Fixed disk adoption breakage
190 - Fixed wrong headers in instance listing for field aliases
195 - Fixed corner case in KVM handling of NICs
196 - Fixed many cases of wrong handling of non-vm_capable nodes
197 - Fixed a bug where a missing instance symlink was not possible to
198 recreate with any ``gnt-*`` command (now ``gnt-instance
199 activate-disks`` does it)
200 - Fixed the volume group name as reported by ``gnt-cluster
202 - Increased timeouts for the import-export code, hopefully leading to
203 fewer aborts due network or instance timeouts
204 - Fixed bug in ``gnt-node list-storage``
205 - Fixed bug where not all daemons were started on cluster
206 initialisation, but only at the first watcher run
207 - Fixed many bugs in the OOB implementation
208 - Fixed watcher behaviour in presence of instances with offline
210 - Fixed instance list output for instances running on the wrong node
211 - a few fixes to the cluster-merge tool, but it still cannot merge
212 multi-node groups (currently it is not recommended to use this tool)
218 - Improved network configuration for the KVM hypervisor
219 - Added e1000 as a supported NIC for Xen-HVM
220 - Improved the lvmstrap tool to also be able to use partitions, as
221 opposed to full disks
222 - Improved speed of disk wiping (the cluster parameter
223 ``prealloc_wipe_disks``, so that it has a low impact on the total time
224 of instance creations
225 - Added documentation for the OS parameters
226 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
227 hypervisor is not responding
228 - Added display of blacklisted and hidden OS information in
230 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
231 NIC and node parameters, which might create problems with currently
232 invalid (but undetected) configuration files, but prevents validation
233 failures when unrelated parameters are modified
234 - Changed cluster initialisation to wait for the master daemon to become
236 - Expanded the RAPI interface:
238 - Added config redistribution resource
239 - Added activation/deactivation of instance disks
240 - Added export of console information
242 - Implemented log file reopening on SIGHUP, which allows using
243 logrotate(8) for the Ganeti log files
244 - Added a basic OOB helper script as an example
250 *(Released Fri, 14 Jan 2011)*
255 - Fixed timezone issues when formatting timestamps
256 - Added support for node groups, available via ``gnt-group`` and other
258 - Added out-of-band framework and management, see :doc:`design
259 document <design-oob>`
260 - Removed support for roman numbers from ``gnt-node list`` and
261 ``gnt-instance list``.
262 - Allowed modification of master network interface via ``gnt-cluster
263 modify --master-netdev``
264 - Accept offline secondaries while shutting down instance disks
265 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
266 - Added support for multiple LVM volume groups
267 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
269 - Added commands to list available fields:
271 - ``gnt-node list-fields``
272 - ``gnt-group list-fields``
273 - ``gnt-instance list-fields``
275 - Updated documentation and man pages
280 - Moved ``rapi_users`` file into separate directory, now named
281 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
283 - Added new tool for running commands on many machines,
284 ``tools/ganeti-listrunner``
285 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
286 improving the ``gnt-instance console`` output
287 - Allowed customisation of disk index separator at ``configure`` time
288 - Export node group allocation policy to :doc:`iallocator <iallocator>`
289 - Added support for non-partitioned md disks in ``lvmstrap``
290 - Added script to gracefully power off KVM instances
291 - Split ``utils`` module into smaller parts
292 - Changed query operations to return more detailed information, e.g.
293 whether an information is unavailable due to an offline node. To use
294 this new functionality, the LUXI call ``Query`` must be used. Field
295 information is now stored by the master daemon and can be retrieved
296 using ``QueryFields``. Instances, nodes and groups can also be queried
297 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
298 exposed via RAPI). The following commands make use of this
299 infrastructure change:
302 - ``gnt-group list-fields``
304 - ``gnt-node list-fields``
305 - ``gnt-instance list``
306 - ``gnt-instance list-fields``
307 - ``gnt-debug locks``
312 - New RAPI resources (see :doc:`rapi`):
316 - ``/2/groups/[group_name]``
317 - ``/2/groups/[group_name]/assign-nodes``
318 - ``/2/groups/[group_name]/modify``
319 - ``/2/groups/[group_name]/rename``
320 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
324 - Implemented ``no_install`` for instance creation
325 - Implemented OS parameters for instance reinstallation, allowing
326 use of special settings on reinstallation (e.g. for preserving data)
331 - Added IPv6 support in import/export
332 - Pause DRBD synchronization while wiping disks on instance creation
333 - Updated unittests and QA scripts
334 - Improved network parameters passed to KVM
335 - Converted man pages from docbook to reStructuredText
341 *(Released Mon, 20 Dec 2010)*
343 Released version 2.3.1~rc1 without any changes.
349 *(Released Wed, 1 Dec 2010)*
351 - impexpd: Disable OpenSSL compression in socat if possible (backport
352 from master, commit e90739d625b, see :doc:`installation guide
353 <install-quick>` for details)
354 - Changed unittest coverage report to exclude test scripts
355 - Added script to check version format
361 *(Released Wed, 1 Dec 2010)*
363 Released version 2.3.0~rc1 without any changes.
369 *(Released Fri, 19 Nov 2010)*
371 A number of bugfixes and documentation updates:
373 - Update ganeti-os-interface documentation
374 - Fixed a bug related to duplicate MACs or similar items which should be
376 - Fix breakage in OS state modify
377 - Reinstall instance: disallow offline secondaries (fixes bug related to
378 OS changing but reinstall failing)
379 - plus all the other fixes between 2.2.1 and 2.2.2
385 *(Released Tue, 2 Nov 2010)*
387 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
388 - Fixed master failover race with watcher
389 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
391 - Accept previously stopped instance for export with instance removal
392 - Simplify and extend the environment variables for instance OS scripts
393 - Added new node flags, ``master_capable`` and ``vm_capable``
394 - Added optional instance disk wiping prior during allocation. This is a
395 cluster-wide option and can be set/modified using
396 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
397 - Added IPv6 support, see :doc:`design document <design-2.3>` and
399 - Added a new watcher option (``--ignore-pause``)
400 - Added option to ignore offline node on instance start/stop
401 (``--ignore-offline``)
402 - Allow overriding OS parameters with ``gnt-instance reinstall``
403 - Added ability to change node's secondary IP address using ``gnt-node
405 - Implemented privilege separation for all daemons except
406 ``ganeti-noded``, see ``configure`` options
407 - Complain if an instance's disk is marked faulty in ``gnt-cluster
409 - Implemented job priorities (see ``ganeti(7)`` manpage)
410 - Ignore failures while shutting down instances during failover from
412 - Exit daemon's bootstrap process only once daemon is ready
413 - Export more information via ``LUInstanceQuery``/remote API
414 - Improved documentation, QA and unittests
415 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
416 restart if the file was created or changed
417 - Added LUXI protocol version sent with each request and response,
418 allowing detection of server/client mismatches
419 - Moved the Python scripts among gnt-* and ganeti-* into modules
420 - Moved all code related to setting up SSH to an external script,
422 - Infrastructure changes for node group support in future versions
428 *(Released Fri, 19 Nov 2010)*
430 A few small bugs fixed, and some improvements to the build system:
432 - Fix documentation regarding conversion to drbd
433 - Fix validation of parameters in cluster modify (``gnt-cluster modify
435 - Fix error handling in node modify with multiple changes
436 - Allow remote imports without checked names
442 *(Released Tue, 19 Oct 2010)*
444 - Disable SSL session ID cache in RPC client
450 *(Released Thu, 14 Oct 2010)*
452 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
453 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
455 - Fix problems with interaction between Curl and Python's HTTP server,
456 resulting in increased speed in many RPC calls
457 - Improve our release script to prevent breakage with older aclocal and
464 *(Released Thu, 7 Oct 2010)*
466 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
467 value retrieved from master
468 - Added support for blacklisted or hidden OS definitions
469 - Added simple lock monitor (accessible via (``gnt-debug locks``)
470 - Added support for -mem-path in KVM hypervisor abstraction layer
471 - Allow overriding instance parameters in tool for inter-cluster
472 instance moves (``tools/move-instance``)
473 - Improved opcode summaries (e.g. in ``gnt-job list``)
474 - Improve consistency of OS listing by sorting it
475 - Documentation updates
481 *(Released Fri, 8 Oct 2010)*
483 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
489 *(Released Mon, 4 Oct 2010)*
491 - Fixed regression in ``gnt-instance rename``
497 *(Released Wed, 22 Sep 2010)*
499 - Fixed OS_VARIANT variable for OS scripts
500 - Fixed cluster tag operations via RAPI
501 - Made ``setup-ssh`` exit with non-zero code if an error occurred
502 - Disabled RAPI CA checks in watcher
508 *(Released Mon, 23 Aug 2010)*
510 - Support DRBD versions of the format "a.b.c.d"
512 - Re-introduce support for usage from multiple threads in RAPI client
513 - Instance renames and modify via RAPI
514 - Work around race condition between processing and archival in job
516 - Mark opcodes following failed one as failed, too
517 - Job field ``lock_status`` was removed due to difficulties making it
518 work with the changed job queue in Ganeti 2.2; a better way to monitor
519 locks is expected for a later 2.2.x release
520 - Fixed dry-run behaviour with many commands
521 - Support ``ssh-agent`` again when adding nodes
522 - Many additional bugfixes
528 *(Released Fri, 30 Jul 2010)*
530 Important change: the internal RPC mechanism between Ganeti nodes has
531 changed from using a home-grown http library (based on the Python base
532 libraries) to use the PycURL library. This requires that PycURL is
533 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
534 against GnuTLS by default. cURL's support for GnuTLS had known issues
535 before cURL 7.21.0 and we recommend using the latest cURL release or
536 linking against OpenSSL. Most other distributions already link PycURL
537 and cURL against OpenSSL. The command::
539 python -c 'import pycurl; print pycurl.version'
541 can be used to determine the libraries PycURL and cURL are linked
544 Other significant changes:
546 - Rewrote much of the internals of the job queue, in order to achieve
547 better parallelism; this decouples job query operations from the job
548 processing, and it should allow much nicer behaviour of the master
549 daemon under load, and it also has uncovered some long-standing bugs
550 related to the job serialisation (now fixed)
551 - Added a default iallocator setting to the cluster parameters,
552 eliminating the need to always pass nodes or an iallocator for
553 operations that require selection of new node(s)
554 - Added experimental support for the LXC virtualization method
555 - Added support for OS parameters, which allows the installation of
556 instances to pass parameter to OS scripts in order to customise the
558 - Added a hypervisor parameter controlling the migration type (live or
559 non-live), since hypervisors have various levels of reliability; this
560 has renamed the 'live' parameter to 'mode'
561 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
562 logical volumes, meaning that cluster verify will ignore them and not
563 flag their presence as errors
564 - The watcher will now reset the error count for failed instances after
565 8 hours, thus allowing self-healing if the problem that caused the
566 instances to be down/fail to start has cleared in the meantime
567 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
568 check for, and warn, if the drbd module parameter ``usermode_helper``
569 is not consistent with the cluster-wide setting; this is needed to
570 make diagnose easier of failed drbd creations
571 - Started adding base IPv6 support, but this is not yet
572 enabled/available for use
573 - Rename operations (cluster, instance) will now return the new name,
574 which is especially useful if a short name was passed in
575 - Added support for instance migration in RAPI
576 - Added a tool to pre-configure nodes for the SSH setup, before joining
577 them to the cluster; this will allow in the future a simplified model
578 for node joining (but not yet fully enabled in 2.2); this needs the
579 paramiko python library
580 - Fixed handling of name-resolving errors
581 - Fixed consistency of job results on the error path
582 - Fixed master-failover race condition when executed multiple times in
584 - Fixed many bugs related to the job queue (mostly introduced during the
585 2.2 development cycle, so not all are impacting 2.1)
586 - Fixed instance migration with missing disk symlinks
587 - Fixed handling of unknown jobs in ``gnt-job archive``
588 - And many other small fixes/improvements
592 - Enhanced both the unittest and the QA coverage
593 - Switched the opcode validation to a generic model, and extended the
594 validation to all opcode parameters
595 - Changed more parts of the code that write shell scripts to use the
597 - Switched the master daemon to use the asyncore library for the Luxi
604 *(Released Thu, 17 Jun 2010)*
606 - Added tool (``move-instance``) and infrastructure to move instances
607 between separate clusters (see :doc:`separate documentation
608 <move-instance>` and :doc:`design document <design-2.2>`)
609 - Added per-request RPC timeout
610 - RAPI now requires a Content-Type header for requests with a body (e.g.
611 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
612 :rfc:`2616` (HTTP/1.1), section 7.2.1)
613 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
615 - Implemented initial support for running Ganeti daemons as separate
616 users, see configure-time flags ``--with-user-prefix`` and
617 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
619 - Instances can be removed after export (``gnt-backup export
621 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
622 key (instead of 1024 bit)
623 - Added new cluster configuration file for cluster domain secret
624 - Import/export now use SSL instead of SSH
625 - Added support for showing estimated time when exporting an instance,
626 see the ``ganeti-os-interface(7)`` manpage and look for
633 *(Released Tue, 16 Nov 2010)*
635 Some more bugfixes. Unless critical bugs occur, this will be the last
638 - Fix case of MAC special-values
639 - Fix mac checker regex
640 - backend: Fix typo causing "out of range" error
641 - Add missing --units in gnt-instance list man page
647 *(Released Tue, 24 Aug 2010)*
650 - Don't ignore secondary node silently on non-mirrored disk templates
652 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
653 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
654 - Properly document the usb_mouse parameter
655 - Fix path in ganeti-rapi(8) (issue 116)
656 - Adjust error message when the ganeti user's .ssh directory is
658 - Add same-node-check when changing the disk template to drbd
664 *(Released Fri, 16 Jul 2010)*
667 - Add an option to only select some reboot types during qa/burnin.
668 (on some hypervisors consequent reboots are not supported)
669 - Fix infrequent race condition in master failover. Sometimes the old
670 master ip address would be still detected as up for a short time
671 after it was removed, causing failover to fail.
672 - Decrease mlockall warnings when the ctypes module is missing. On
673 Python 2.4 we support running even if no ctypes module is installed,
674 but we were too verbose about this issue.
675 - Fix building on old distributions, on which man doesn't have a
677 - Fix RAPI not to ignore the MAC address on instance creation
678 - Implement the old instance creation format in the RAPI client.
684 *(Released Thu, 01 Jul 2010)*
686 A small bugfix release:
687 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
688 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
689 a non-existing option
690 - Fix instance create: the --force-variant option was ignored
691 - Improve pylint 0.21 compatibility and warnings with Python 2.6
692 - Fix modify node storage with non-FQDN arguments
693 - Fix RAPI client to authenticate under Python 2.6 when used
694 for more than 5 requests needing authentication
695 - Fix gnt-instance modify -t (storage) giving a wrong error message
696 when converting a non-shutdown drbd instance to plain
702 *(Released Fri, 18 Jun 2010)*
704 A small bugfix release:
706 - Fix live migration of KVM instances started with older Ganeti
707 versions which had fewer hypervisor parameters
708 - Fix gnt-instance grow-disk on down instances
709 - Fix an error-reporting bug during instance migration
710 - Better checking of the ``--net`` and ``--disk`` values, to avoid
711 silently ignoring broken ones
712 - Fix an RPC error reporting bug affecting, for example, RAPI client
714 - Fix bug triggered by different API version os-es on different nodes
715 - Fix a bug in instance startup with custom hvparams: OS level
716 parameters would fail to be applied.
717 - Fix the RAPI client under Python 2.6 (but more work is needed to
718 make it work completely well with OpenSSL)
719 - Fix handling of errors when resolving names from DNS
725 *(Released Thu, 3 Jun 2010)*
727 A medium sized development cycle. Some new features, and some
728 fixes/small improvements/cleanups.
733 The node deamon now tries to mlock itself into memory, unless the
734 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
735 its logs, and falls back to console logging. This allows emergency
736 features such as ``gnt-node powercycle`` to work even in the event of a
737 broken node disk (tested offlining the disk hosting the node's
738 filesystem and dropping its memory caches; don't try this at home)
740 KVM: add vhost-net acceleration support. It can be tested with a new
741 enough version of the kernel and of qemu-kvm.
743 KVM: Add instance chrooting feature. If you use privilege dropping for
744 your VMs you can also now force them to chroot to an empty directory,
745 before starting the emulated guest.
747 KVM: Add maximum migration bandwith and maximum downtime tweaking
748 support (requires a new-enough version of qemu-kvm).
750 Cluster verify will now warn if the master node doesn't have the master
753 Add a new (incompatible) instance creation request format to RAPI which
754 supports all parameters (previously only a subset was supported, and it
755 wasn't possible to extend the old format to accomodate all the new
756 features. The old format is still supported, and a client can check for
757 this feature, before using it, by checking for its presence in the
758 ``features`` RAPI resource.
760 Now with ancient latin support. Try it passing the ``--roman`` option to
761 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
762 (requires the python-roman module to be installed, in order to work).
767 As usual many internal code refactorings, documentation updates, and
770 - Lots of improvements and cleanups to the experimental Remote API
771 (RAPI) client library.
772 - A new unit test suite for the core daemon libraries.
773 - A fix to creating missing directories makes sure the umask is not
774 applied anymore. This enforces the same directory permissions
776 - Better handling terminating daemons with ctrl+c (used when running
777 them in debugging mode).
778 - Fix a race condition in live migrating a KVM instance, when stat()
779 on the old proc status file returned EINVAL, which is an unexpected
781 - Fixed manpage checking with newer man and utf-8 charachters. But now
782 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
788 *(Released Fri, 7 May 2010)*
790 Fix a bug which prevented untagged KVM instances from starting.
796 *(Released Fri, 7 May 2010)*
798 Another release with a long development cycle, during which many
799 different features were added.
804 The KVM hypervisor now can run the individual instances as non-root, to
805 reduce the impact of a VM being hijacked due to bugs in the
806 hypervisor. It is possible to run all instances as a single (non-root)
807 user, to manually specify a user for each instance, or to dynamically
808 allocate a user out of a cluster-wide pool to each instance, with the
809 guarantee that no two instances will run under the same user ID on any
812 An experimental RAPI client library, that can be used standalone
813 (without the other Ganeti libraries), is provided in the source tree as
814 ``lib/rapi/client.py``. Note this client might change its interface in
815 the future, as we iterate on its capabilities.
817 A new command, ``gnt-cluster renew-crypto`` has been added to easily
818 replace the cluster's certificates and crypto keys. This might help in
819 case they have been compromised, or have simply expired.
821 A new disk option for instance creation has been added that allows one
822 to "adopt" currently existing logical volumes, with data
823 preservation. This should allow easier migration to Ganeti from
824 unmanaged (or managed via other software) instances.
826 Another disk improvement is the possibility to convert between redundant
827 (DRBD) and plain (LVM) disk configuration for an instance. This should
828 allow better scalability (starting with one node and growing the
829 cluster, or shrinking a two-node cluster to one node).
831 A new feature that could help with automated node failovers has been
832 implemented: if a node sees itself as offline (by querying the master
833 candidates), it will try to shutdown (hard) all instances and any active
834 DRBD devices. This reduces the risk of duplicate instances if an
835 external script automatically failovers the instances on such nodes. To
836 enable this, the cluster parameter ``maintain_node_health`` should be
837 enabled; in the future this option (per the name) will enable other
838 automatic maintenance features.
840 Instance export/import now will reuse the original instance
841 specifications for all parameters; that means exporting an instance,
842 deleting it and the importing it back should give an almost identical
843 instance. Note that the default import behaviour has changed from
844 before, where it created only one NIC; now it recreates the original
847 Cluster verify has added a few new checks: SSL certificates validity,
848 /etc/hosts consistency across the cluster, etc.
853 As usual, many internal changes were done, documentation fixes,
856 - Fixed cluster initialization with disabled cluster storage (regression
858 - File-based storage supports growing the disks
859 - Fixed behaviour of node role changes
860 - Fixed cluster verify for some corner cases, plus a general rewrite of
861 cluster verify to allow future extension with more checks
862 - Fixed log spamming by watcher and node daemon (regression introduced
864 - Fixed possible validation issues when changing the list of enabled
866 - Fixed cleanup of /etc/hosts during node removal
867 - Fixed RAPI response for invalid methods
868 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
869 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
870 from ide disks, etc.)
871 - Allow OS changes without re-installation (to record a changed OS
872 outside of Ganeti, or to allow OS renames)
873 - Allow instance creation without OS installation (useful for example if
874 the OS will be installed manually, or restored from a backup not in
876 - Implemented option to make cluster ``copyfile`` use the replication
878 - Added list of enabled hypervisors to ssconf (possibly useful for
880 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
882 - A partial form of node re-IP is possible via node readd, which now
883 allows changed node primary IP
884 - Command line utilities now show an informational message if the job is
886 - The logs of the master daemon now show the PID/UID/GID of the
893 *(Released Fri, 12 Mar 2010)*
895 During the 2.1.0 long release candidate cycle, a lot of improvements and
896 changes have accumulated with were released later as 2.1.1.
901 The node evacuate command (``gnt-node evacuate``) was significantly
902 rewritten, and as such the IAllocator protocol was changed - a new
903 request type has been added. This unfortunate change during a stable
904 series is designed to improve performance of node evacuations; on
905 clusters with more than about five nodes and which are well-balanced,
906 evacuation should proceed in parallel for all instances of the node
907 being evacuated. As such, any existing IAllocator scripts need to be
908 updated, otherwise the above command will fail due to the unknown
909 request. The provided "dumb" allocator has not been updated; but the
910 ganeti-htools package supports the new protocol since version 0.2.4.
912 Another important change is increased validation of node and instance
913 names. This might create problems in special cases, if invalid host
914 names are being used.
916 Also, a new layer of hypervisor parameters has been added, that sits at
917 OS level between the cluster defaults and the instance ones. This allows
918 customisation of virtualization parameters depending on the installed
919 OS. For example instances with OS 'X' may have a different KVM kernel
920 (or any other parameter) than the cluster defaults. This is intended to
921 help managing a multiple OSes on the same cluster, without manual
922 modification of each instance's parameters.
924 A tool for merging clusters, ``cluster-merge``, has been added in the
930 - Improved the int/float conversions that should make the code more
931 robust in face of errors from the node daemons
932 - Fixed the remove node code in case of internal configuration errors
933 - Fixed the node daemon behaviour in face of inconsistent queue
934 directory (e.g. read-only file-system where we can't open the files
936 - Fixed the behaviour of gnt-node modify for master candidate demotion;
937 now it either aborts cleanly or, if given the new "auto_promote"
938 parameter, will automatically promote other nodes as needed
939 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
940 completely prevent Ganeti from working
941 - Fixed bug for instance export when not all disks were successfully
943 - Fixed behaviour of node add when the new node is slow in starting up
945 - Fixed handling of signals in the LUXI client, which should improve
946 behaviour of command-line scripts
947 - Added checks for invalid node/instance names in the configuration (now
948 flagged during cluster verify)
949 - Fixed watcher behaviour for disk activation errors
950 - Fixed two potentially endless loops in http library, which led to the
951 RAPI daemon hanging and consuming 100% CPU in some cases
952 - Fixed bug in RAPI daemon related to hashed passwords
953 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
955 - Enhanced compatibility with non-Debian OSes, but not using absolute
956 path in some commands and allowing customisation of the ssh
957 configuration directory
958 - Fixed possible future issue with new Python versions by abiding to the
959 proper use of ``__slots__`` attribute on classes
960 - Added checks that should prevent directory traversal attacks
961 - Many documentation fixes based on feedback from users
966 - Added an "early_release" more for instance replace disks and node
967 evacuate, where we release locks earlier and thus allow higher
968 parallelism within the cluster
969 - Added watcher hooks, intended to allow the watcher to restart other
970 daemons (e.g. from the ganeti-nbma project), but they can be used of
971 course for any other purpose
972 - Added a compile-time disable for DRBD barriers, to increase
973 performance if the administrator trusts the power supply or the
974 storage system to not lose writes
975 - Added the option of using syslog for logging instead of, or in
976 addition to, Ganeti's own log files
977 - Removed boot restriction for paravirtual NICs for KVM, recent versions
978 can indeed boot from a paravirtual NIC
979 - Added a generic debug level for many operations; while this is not
980 used widely yet, it allows one to pass the debug value all the way to
982 - Enhanced the hooks environment for instance moves (failovers,
983 migrations) where the primary/secondary nodes changed during the
984 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
985 - Enhanced data validations for many user-supplied values; one important
986 item is the restrictions imposed on instance and node names, which
987 might reject some (invalid) host names
988 - Add a configure-time option to disable file-based storage, if it's not
989 needed; this allows greater security separation between the master
990 node and the other nodes from the point of view of the inter-node RPC
992 - Added user notification in interactive tools if job is waiting in the
993 job queue or trying to acquire locks
994 - Added log messages when a job is waiting for locks
995 - Added filtering by node tags in instance operations which admit
996 multiple instances (start, stop, reboot, reinstall)
997 - Added a new tool for cluster mergers, ``cluster-merge``
998 - Parameters from command line which are of the form ``a=b,c=d`` can now
999 use backslash escapes to pass in values which contain commas,
1000 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1002 - For KVM, the instance name is the first parameter passed to KVM, so
1003 that it's more visible in the process list
1009 *(Released Tue, 2 Mar 2010)*
1011 Ganeti 2.1 brings many improvements with it. Major changes:
1013 - Added infrastructure to ease automated disk repairs
1014 - Added new daemon to export configuration data in a cheaper way than
1015 using the remote API
1016 - Instance NICs can now be routed instead of being associated with a
1018 - Improved job locking logic to reduce impact of jobs acquiring multiple
1019 locks waiting for other long-running jobs
1021 In-depth implementation details can be found in the Ganeti 2.1 design
1027 - Added chroot hypervisor
1028 - Added more options to xen-hvm hypervisor (``kernel_path`` and
1030 - Added more options to xen-pvm hypervisor (``use_bootloader``,
1031 ``bootloader_path`` and ``bootloader_args``)
1032 - Added the ``use_localtime`` option for the xen-hvm and kvm
1033 hypervisors, and the default value for this has changed to false (in
1034 2.0 xen-hvm always enabled it)
1035 - Added luxi call to submit multiple jobs in one go
1036 - Added cluster initialization option to not modify ``/etc/hosts``
1038 - Added network interface parameters
1039 - Added dry run mode to some LUs
1040 - Added RAPI resources:
1042 - ``/2/instances/[instance_name]/info``
1043 - ``/2/instances/[instance_name]/replace-disks``
1044 - ``/2/nodes/[node_name]/evacuate``
1045 - ``/2/nodes/[node_name]/migrate``
1046 - ``/2/nodes/[node_name]/role``
1047 - ``/2/nodes/[node_name]/storage``
1048 - ``/2/nodes/[node_name]/storage/modify``
1049 - ``/2/nodes/[node_name]/storage/repair``
1051 - Added OpCodes to evacuate or migrate all instances on a node
1052 - Added new command to list storage elements on nodes (``gnt-node
1053 list-storage``) and modify them (``gnt-node modify-storage``)
1054 - Added new ssconf files with master candidate IP address
1055 (``ssconf_master_candidates_ips``), node primary IP address
1056 (``ssconf_node_primary_ips``) and node secondary IP address
1057 (``ssconf_node_secondary_ips``)
1058 - Added ``ganeti-confd`` and a client library to query the Ganeti
1059 configuration via UDP
1060 - Added ability to run hooks after cluster initialization and before
1062 - Added automatic mode for disk replace (``gnt-instance replace-disks
1064 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1065 after catastrophic data-loss
1066 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
1068 - Added ``gnt-instance move`` command to move instances
1069 - Added ``gnt-cluster watcher`` command to control watcher
1070 - Added ``gnt-node powercycle`` command to powercycle nodes
1071 - Added new job status field ``lock_status``
1072 - Added parseable error codes to cluster verification (``gnt-cluster
1073 verify --error-codes``) and made output less verbose (use
1074 ``--verbose`` to restore previous behaviour)
1075 - Added UUIDs to the main config entities (cluster, nodes, instances)
1076 - Added support for OS variants
1077 - Added support for hashed passwords in the Ganeti remote API users file
1079 - Added option to specify maximum timeout on instance shutdown
1080 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
1081 - Added new helper script to start and stop Ganeti daemons
1082 (``daemon-util``), with the intent to reduce the work necessary to
1083 adjust Ganeti for non-Debian distributions and to start/stop daemons
1085 - Added more unittests
1086 - Fixed critical bug in ganeti-masterd startup
1087 - Removed the configure-time ``kvm-migration-port`` parameter, this is
1088 now customisable at the cluster level for both the KVM and Xen
1089 hypervisors using the new ``migration_port`` parameter
1090 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
1091 reinstalling an instance
1092 - Allowed ``@`` in tag names
1093 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
1094 - Many documentation updates
1095 - Distribute hypervisor files on ``gnt-cluster redist-conf``
1096 - ``gnt-instance reinstall`` can now reinstall multiple instances
1097 - Updated many command line parameters
1098 - Introduced new OS API version 15
1099 - No longer support a default hypervisor
1100 - Treat virtual LVs as inexistent
1101 - Improved job locking logic to reduce lock contention
1102 - Match instance and node names case insensitively
1103 - Reimplemented bash completion script to be more complete
1110 *(Released Thu, 4 Feb 2010)*
1112 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
1113 - Fix a string formatting bug
1114 - Improve safety of the code in some error paths
1115 - Improve data validation in the master of values returned from nodes
1121 *(Released Thu, 17 Dec 2009)*
1123 - Fix security issue due to missing validation of iallocator names; this
1124 allows local and remote execution of arbitrary executables
1125 - Fix failure of gnt-node list during instance removal
1126 - Ship the RAPI documentation in the archive
1132 *(Released Wed, 30 Sep 2009)*
1134 - Fixed many wrong messages
1135 - Fixed a few bugs related to the locking library
1136 - Fixed MAC checking at instance creation time
1137 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
1138 - Fixed a few issues related to signal handling in both daemons and
1140 - Fixed the example startup script provided
1141 - Fixed insserv dependencies in the example startup script (patch from
1143 - Fixed handling of drained nodes in the iallocator framework
1144 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
1146 - Fixed error related to invalid job IDs in job polling
1147 - Fixed job/opcode persistence on unclean master shutdown
1148 - Fixed handling of partial job processing after unclean master
1150 - Fixed error reporting from LUs, previously all errors were converted
1151 into execution errors
1152 - Fixed error reporting from burnin
1153 - Decreased significantly the memory usage of the job queue
1154 - Optimised slightly multi-job submission
1155 - Optimised slightly opcode loading
1156 - Backported the multi-job submit framework from the development
1157 branch; multi-instance start and stop should be faster
1158 - Added script to clean archived jobs after 21 days; this will reduce
1159 the size of the queue directory
1160 - Added some extra checks in disk size tracking
1161 - Added an example ethers hook script
1162 - Added a cluster parameter that prevents Ganeti from modifying of
1164 - Added more node information to RAPI responses
1165 - Added a ``gnt-job watch`` command that allows following the ouput of a
1167 - Added a bind-address option to ganeti-rapi
1168 - Added more checks to the configuration verify
1169 - Enhanced the burnin script such that some operations can be retried
1171 - Converted instance reinstall to multi-instance model
1177 *(Released Fri, 7 Aug 2009)*
1179 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
1180 to allow using the pre-2.0.2 behaviour in activation, if any existing
1181 instances have mismatched disk sizes in the configuration
1182 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
1183 any configuration mismatches for disk sizes
1184 - Added ``gnt-master cluste-failover --no-voting`` to allow master
1185 failover to work on two-node clusters
1186 - Fixed the ``--net`` option of ``gnt-backup import``, which was
1188 - Fixed detection of OS script errors in ``gnt-backup export``
1189 - Fixed exit code of ``gnt-backup export``
1195 *(Released Fri, 17 Jul 2009)*
1197 - Added experimental support for stripped logical volumes; this should
1198 enhance performance but comes with a higher complexity in the block
1199 device handling; stripping is only enabled when passing
1200 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
1201 affected even in the non-stripped mode
1202 - Improved resiliency against transient failures at the end of DRBD
1203 resyncs, and in general of DRBD resync checks
1204 - Fixed a couple of issues with exports and snapshot errors
1205 - Fixed a couple of issues in instance listing
1206 - Added display of the disk size in ``gnt-instance info``
1207 - Fixed checking for valid OSes in instance creation
1208 - Fixed handling of the "vcpus" parameter in instance listing and in
1209 general of invalid parameters
1210 - Fixed http server library, and thus RAPI, to handle invalid
1211 username/password combinations correctly; this means that now they
1212 report unauthorized for queries too, not only for modifications,
1213 allowing earlier detect of configuration problems
1214 - Added a new "role" node list field, equivalent to the master/master
1215 candidate/drained/offline flags combinations
1216 - Fixed cluster modify and changes of candidate pool size
1217 - Fixed cluster verify error messages for wrong files on regular nodes
1218 - Fixed a couple of issues with node demotion from master candidate role
1219 - Fixed node readd issues
1220 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
1221 - Added a new ``--no-voting`` option for masterfailover to fix failover
1222 on two-nodes clusters when the former master node is unreachable
1223 - Added instance reinstall over RAPI
1229 *(Released Tue, 16 Jun 2009)*
1231 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
1232 allow re-adding the start in single-user option (regression from 1.2)
1233 - the watcher writes the instance status to a file, to allow monitoring
1234 to report the instance status (from the master) based on cached
1235 results of the watcher's queries; while this can get stale if the
1236 watcher is being locked due to other work on the cluster, this is
1237 still an improvement
1238 - the watcher now also restarts the node daemon and the rapi daemon if
1240 - fixed the watcher to handle full and drained queue cases
1241 - hooks export more instance data in the environment, which helps if
1242 hook scripts need to take action based on the instance's properties
1243 (no longer need to query back into ganeti)
1244 - instance failovers when the instance is stopped do not check for free
1245 RAM, so that failing over a stopped instance is possible in low memory
1247 - rapi uses queries for tags instead of jobs (for less job traffic), and
1248 for cluster tags it won't talk to masterd at all but read them from
1250 - a couple of error handling fixes in RAPI
1251 - drbd handling: improved the error handling of inconsistent disks after
1252 resync to reduce the frequency of "there are some degraded disks for
1253 this instance" messages
1254 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
1255 error handling path called a wrong function name)
1261 *(Released Wed, 27 May 2009)*
1263 - no changes from rc5
1266 Version 2.0 release candidate 5
1267 -------------------------------
1269 *(Released Wed, 20 May 2009)*
1271 - fix a couple of bugs (validation, argument checks)
1272 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
1273 - some small improvements to RAPI and IAllocator
1274 - make watcher automatically start the master daemon if down
1277 Version 2.0 release candidate 4
1278 -------------------------------
1280 *(Released Mon, 27 Apr 2009)*
1282 - change the OS list to not require locks; this helps with big clusters
1283 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
1284 volume group is broken
1285 - ``gnt-instance info``, without any arguments, doesn't run for all
1286 instances anymore; either pass ``--all`` or pass the desired
1287 instances; this helps against mistakes on big clusters where listing
1288 the information for all instances takes a long time
1289 - miscellaneous doc and man pages fixes
1292 Version 2.0 release candidate 3
1293 -------------------------------
1295 *(Released Wed, 8 Apr 2009)*
1297 - Change the internal locking model of some ``gnt-node`` commands, in
1298 order to reduce contention (and blocking of master daemon) when
1299 batching many creation/reinstall jobs
1300 - Fixes to Xen soft reboot
1301 - No longer build documentation at build time, instead distribute it in
1302 the archive, in order to reduce the need for the whole docbook/rst
1306 Version 2.0 release candidate 2
1307 -------------------------------
1309 *(Released Fri, 27 Mar 2009)*
1311 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
1312 - Fix watcher startup sequence, improves the behaviour of busy clusters
1313 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
1314 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
1315 bind address and other places
1316 - Some documentation fixes and updates
1319 Version 2.0 release candidate 1
1320 -------------------------------
1322 *(Released Mon, 2 Mar 2009)*
1324 - More documentation updates, now all docs should be more-or-less
1326 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
1328 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
1330 - Fix an issue related to $libdir/run/ganeti and cluster creation
1336 *(Released Thu, 19 Feb 2009)*
1338 - Xen PVM and KVM have switched the default value for the instance root
1339 disk to the first partition on the first drive, instead of the whole
1340 drive; this means that the OS installation scripts must be changed
1342 - Man pages have been updated
1343 - RAPI has been switched by default to HTTPS, and the exported functions
1344 should all work correctly
1345 - RAPI v1 has been removed
1346 - Many improvements to the KVM hypervisor
1347 - Block device errors are now better reported
1348 - Many other bugfixes and small improvements
1354 *(Released Mon, 26 Jan 2009)*
1356 - Version 2 is a general rewrite of the code and therefore the
1357 differences are too many to list, see the design document for 2.0 in
1358 the ``doc/`` subdirectory for more details
1359 - In this beta version there is not yet a migration path from 1.2 (there
1360 will be one in the final 2.0 release)
1361 - A few significant changes are:
1363 - all commands are executed by a daemon (``ganeti-masterd``) and the
1364 various ``gnt-*`` commands are just front-ends to it
1365 - all the commands are entered into, and executed from a job queue,
1366 see the ``gnt-job(8)`` manpage
1367 - the RAPI daemon supports read-write operations, secured by basic
1368 HTTP authentication on top of HTTPS
1369 - DRBD version 0.7 support has been removed, DRBD 8 is the only
1370 supported version (when migrating from Ganeti 1.2 to 2.0, you need
1371 to migrate to DRBD 8 first while still running Ganeti 1.2)
1372 - DRBD devices are using statically allocated minor numbers, which
1373 will be assigned to existing instances during the migration process
1374 - there is support for both Xen PVM and Xen HVM instances running on
1376 - KVM virtualization is supported too
1377 - file-based storage has been implemented, which means that it is
1378 possible to run the cluster without LVM and DRBD storage, for
1379 example using a shared filesystem exported from shared storage (and
1380 still have live migration)
1386 *(Released Tue, 13 Jan 2009)*
1388 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
1389 - Reuse the old instance mac address by default on instance import, if
1390 the instance name is the same.
1391 - Handle situations in which the node info rpc returns incomplete
1393 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
1394 - Improved version of batcher:
1396 - state file support
1397 - instance mac address support
1398 - support for HVM clusters/instances
1400 - Add an option to show the number of cpu sockets and nodes in
1402 - Support OSes that handle more than one version of the OS api (but do
1403 not change the current API in any other way)
1404 - Fix ``gnt-node migrate``
1405 - ``gnt-debug`` man page
1406 - Fixes various more typos and small issues
1407 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
1413 *(Released Wed, 24 Sep 2008)*
1415 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
1416 type of disk exported to fully virtualized instances.
1417 - provide access to the serial console of HVM instances
1418 - instance auto_balance flag, set by default. If turned off it will
1419 avoid warnings on cluster verify if there is not enough memory to fail
1420 over an instance. in the future it will prevent automatically failing
1421 it over when we will support that.
1422 - batcher tool for instance creation, see ``tools/README.batcher``
1423 - ``gnt-instance reinstall --select-os`` to interactively select a new
1424 operating system when reinstalling an instance.
1425 - when changing the memory amount on instance modify a check has been
1426 added that the instance will be able to start. also warnings are
1427 emitted if the instance will not be able to fail over, if auto_balance
1429 - documentation fixes
1430 - sync fields between ``gnt-instance list/modify/add/import``
1431 - fix a race condition in drbd when the sync speed was set after giving
1432 the device a remote peer.
1438 *(Released Tue, 22 Jul 2008)*
1440 - note: the allowed size and number of tags per object were reduced
1441 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
1442 - fixed twisted 8.x compatibility
1443 - fixed ``gnt-instance replace-disks`` with iallocator
1444 - add TCP keepalives on twisted connections to detect restarted nodes
1445 - disk increase support, see ``gnt-instance grow-disk``
1446 - implement bulk node/instance query for RAPI
1447 - add tags in node/instance listing (optional)
1448 - experimental migration (and live migration) support, read the man page
1449 for ``gnt-instance migrate``
1450 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
1451 has some small improvements in handling its state file
1457 *(Released Fri, 13 Jun 2008)*
1459 - Experimental readonly, REST-based remote API implementation;
1460 automatically started on master node, TCP port 5080, if enabled by
1461 ``--enable-rapi`` parameter to configure script.
1462 - Instance allocator support. Add and import instance accept a
1463 ``--iallocator`` parameter, and call that instance allocator to decide
1464 which node to use for the instance. The iallocator document describes
1465 what's expected from an allocator script.
1466 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
1467 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
1468 node is lost there is still enough memory to fail over the instances
1470 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
1471 ``gnt-cluster verify``, to check for site-specific compliance. All the
1472 hooks will run, and their output, if any, will be displayed. Any
1473 failing hook will make the verification return an error value.
1474 - ``gnt-cluster verify`` now checks that its peers are reachable on the
1475 primary and secondary interfaces
1476 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
1477 that is still declared as part of the cluster and has failed.
1478 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
1479 specifying output fields, that just adds the chosen fields to the
1481 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
1482 export from the filesystem.
1483 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
1484 have been added. Using them you can enable/disable acpi and pae
1485 support, and specify a path for a cd image to be exported to the
1486 instance. These parameters as the name suggest only work on HVM
1488 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
1489 PAE support will be set to the previously hardcoded values, but the
1490 (previously hardcoded) path to the CDROM ISO image will be unset and
1491 if required, needs to be set manually with ``gnt-instance modify``
1493 - The address to which an instance's VNC console is bound is now
1494 selectable per-instance, rather than being cluster wide. Of course
1495 this only applies to instances controlled via VNC, so currently just
1496 applies to HVM clusters.
1502 *(Released Mon, 18 Feb 2008)*
1504 - more tweaks to the disk activation code (especially helpful for DRBD)
1505 - change the default ``gnt-instance list`` output format, now there is
1506 one combined status field (see the manpage for the exact values this
1508 - some more fixes for the mac export to hooks change
1509 - make Ganeti not break with DRBD 8.2.x (which changed the version
1510 format in ``/proc/drbd``) (issue 24)
1511 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
1512 template, allowing migration from DRBD0.7+MD to DRBD8
1518 *(Released Wed, 30 Jan 2008)*
1520 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1522 - add command aliases infrastructure and a few aliases
1523 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1524 man pages and the ``--help`` option of ``gnt-node
1525 list``/``gnt-instance list``
1526 - fix ``gnt-backup list`` with down nodes (issue 21)
1527 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1528 - fix the dist archive and add a check for including svn/git files in
1530 - some developer-related changes: improve the burnin and the QA suite,
1531 add an upload script for testing during development
1537 *(Released Wed, 16 Jan 2008)*
1539 - experimental HVM support, read the install document, section
1540 "Initializing the cluster"
1541 - allow for the PVM hypervisor per-instance kernel and initrd paths
1542 - add a new command ``gnt-cluster verify-disks`` which uses a new
1543 algorithm to improve the reconnection of the DRBD pairs if the device
1544 on the secondary node has gone away
1545 - make logical volume code auto-activate LVs at disk activation time
1546 - slightly improve the speed of activating disks
1547 - allow specification of the MAC address at instance creation time, and
1548 changing it later via ``gnt-instance modify``
1549 - fix handling of external commands that generate lots of output on
1551 - update documentation with regard to minimum version of DRBD8 supported
1557 *(Released Tue, 4 Dec 2007)*
1559 - Log the ``xm create`` output to the node daemon log on failure (to
1560 help diagnosing the error)
1561 - In debug mode, log all external commands output if failed to the logs
1562 - Change parsing of lvm commands to ignore stderr
1568 *(Released Wed, 28 Nov 2007)*
1570 - Another round of updates to the DRBD 8 code to deal with more failures
1571 in the replace secondary node operation
1572 - Some more logging of failures in disk operations (lvm, drbd)
1573 - A few documentation updates
1580 *(Released Tue, 13 Nov 2007)*
1582 - Change configuration file format from Python's Pickle to JSON.
1583 Upgrading is possible using the cfgupgrade utility.
1584 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1585 faster replace disks and is more stable (DRBD 8 has many improvements
1586 compared to DRBD 0.7)
1587 - Added command line tags support (see man pages for ``gnt-instance``,
1588 ``gnt-node``, ``gnt-cluster``)
1589 - Added instance rename support
1590 - Added multi-instance startup/shutdown
1591 - Added cluster rename support
1592 - Added ``gnt-node evacuate`` to simplify some node operations
1593 - Added instance reboot operation that can speedup reboot as compared to
1595 - Soften the requirement that hostnames are in FQDN format
1596 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1598 - Removed dependency on debian's patched fping that uses the
1599 non-standard ``-S`` option
1600 - Now the OS definitions are searched for in multiple, configurable
1601 paths (easier for distros to package)
1602 - Some changes to the hooks infrastructure (especially the new
1603 post-configuration update hook)
1604 - Other small bugfixes
1606 .. vim: set textwidth=72 syntax=rst :