8 *(Released Fri, 31 Jan 2014)*
10 Incompatible/important changes
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
13 - Adding disks with 'gnt-instance modify' now waits for the disks to sync per
14 default. Specify --no-wait-for-sync to override this behavior.
15 - The Ganeti python code now adheres to a private-module layout. In particular,
16 the module 'ganeti' is no longer in the python search path.
17 - On instance allocation, the iallocator now considers non-LVM storage
18 properly. In particular, actual file storage space information is used
19 when allocating space for a file/sharedfile instance.
20 - When disabling disk templates cluster-wide, the cluster now first
21 checks whether there are instances still using those templates.
22 - 'gnt-node list-storage' now also reports storage information about
23 file-based storage types.
24 - In case of non drbd instances, export \*_SECONDARY environment variables
25 as empty strings (and not "None") during 'instance-migrate' related hooks.
30 - KVM hypervisors can now access RBD storage directly without having to
31 go through a block device.
32 - A new command 'gnt-cluster upgrade' was added that automates the upgrade
33 procedure between two Ganeti versions that are both 2.10 or higher.
34 - The move-instance command can now change disk templates when moving
35 instances, and does not require any node placement options to be
36 specified if the destination cluster has a default iallocator.
37 - Users can now change the soundhw and cpuid settings for XEN hypervisors.
38 - Hail and hbal now have the (optional) capability of accessing average CPU
39 load information through the monitoring deamon, and to use it to dynamically
40 adapt the allocation of instances.
41 - Hotplug support. Introduce new option '--hotplug' to ``gnt-instance modify``
42 so that disk and NIC modifications take effect without the need of actual
43 reboot. There are a couple of constrains currently for this feature:
45 - only KVM hypervisor (versions >= 1.0) supports it,
46 - one can not (yet) hotplug a disk using userspace access mode for RBD
47 - in case of a downgrade instances should suffer a reboot in order to
48 be migratable (due to core change of runtime files)
53 - A new test framework for logical units was introduced and the test
54 coverage for logical units was improved significantly.
55 - Opcodes are entirely generated from Haskell using the tool 'hs2py' and
56 the module 'src/Ganeti/OpCodes.hs'.
57 - Constants are also generated from Haskell using the tool
58 'hs2py-constants' and the module 'src/Ganeti/Constants.hs', with the
59 exception of socket related constants, which require changing the
60 cluster configuration file, and HVS related constants, because they
61 are part of a port of instance queries to Haskell. As a result, these
62 changes will be part of the next release of Ganeti.
67 The following new dependencies have been added/updated.
71 - The version requirements for ``python-mock`` have increased to at least
72 version 1.0.1. It is still used for testing only.
77 - Documentation improvements
78 - Run drbdsetup syncer only on network attach
79 - Include target node in hooks nodes for migration
81 - Support post-upgrade hooks during cluster upgrades
83 Inherited from the 2.9 branch:
85 - Ensure that all the hypervisors exist in the config file (Issue 640)
86 - Correctly recognise the role as master node (Issue 687)
87 - configure: allow detection of Sphinx 1.2+ (Issue 502)
88 - gnt-instance now honors the KVM path correctly (Issue 691)
90 Inherited from the 2.8 branch:
92 - Change the list separator for the usb_devices parameter from comma to space.
93 Commas could not work because they are already the hypervisor option
95 - Add support for blktap2 file-driver (Issue 638)
96 - Add network tag definitions to the haskell codebase (Issue 641)
97 - Fix RAPI network tag handling
98 - Add the network tags to the tags searched by gnt-cluster search-tags
99 - Fix caching bug preventing jobs from being cancelled
100 - Start-master/stop-master was always failing if ConfD was disabled. (Issue 685)
105 - All known issues in 2.10.0 beta1 have been resolved (see changes from
107 - Improve handling of KVM runtime files from earlier Ganeti versions
108 - Documentation fixes
110 Inherited from the 2.9 branch:
112 - use custom KVM path if set for version checking
113 - SingleNotifyPipeCondition: don't share pollers
115 Inherited from the 2.8 branch:
117 - Fixed Luxi daemon socket permissions after master-failover
118 - Improve IP version detection code directly checking for colons rather than
119 passing the family from the cluster object
120 - Fix NODE/NODE_RES locking in LUInstanceCreate by not acquiring NODE_RES locks
121 opportunistically anymore (Issue 622)
122 - Allow link local IPv6 gateways (Issue 624)
123 - Fix error printing (Issue 616)
124 - Fix a bug in InstanceSetParams concerning names: in case no name is passed in
125 disk modifications, keep the old one. If name=none then set disk name to
127 - Update build_chroot script to work with the latest hackage packages
128 - Add a packet number limit to "fping" in master-ip-setup (Issue 630)
129 - Fix evacuation out of drained node (Issue 615)
130 - Add default file_driver if missing (Issue 571)
131 - Fix job error message after unclean master shutdown (Issue 618)
132 - Lock group(s) when creating instances (Issue 621)
133 - SetDiskID() before accepting an instance (Issue 633)
134 - Allow the ext template disks to receive arbitrary parameters, both at creation
135 time and while being modified
136 - Xen handle domain shutdown (future proofing cherry-pick)
137 - Refactor reading live data in htools (future proofing cherry-pick)
143 *(Released Tue, 17 Dec 2013)*
145 This was the first RC release of the 2.10 series. All important changes
146 are listed in the latest 2.10 entry.
152 *(Released Wed, 27 Nov 2013)*
154 This was the first beta release of the 2.10 series. All important changes
155 are listed in the latest 2.10 entry.
160 The following issues are known to be present in the beta and will be fixed
163 - Issue 477: Wrong permissions for confd LUXI socket
164 - Issue 621: Instance related opcodes do not aquire network/group locks
165 - Issue 622: Assertion Error: Node locks differ from node resource locks
166 - Issue 623: IPv6 Masterd <-> Luxid communication error
172 *(Released Mon, 27 Jan 2014)*
174 - Ensure that all the hypervisors exist in the config file (Issue 640)
175 - Correctly recognise the role as master node (Issue 687)
176 - configure: allow detection of Sphinx 1.2+ (Issue 502)
177 - gnt-instance now honors the KVM path correctly (Issue 691)
179 Inherited from the 2.8 branch:
181 - Change the list separator for the usb_devices parameter from comma to space.
182 Commas could not work because they are already the hypervisor option
183 separator (Issue 649)
184 - Add support for blktap2 file-driver (Issue 638)
185 - Add network tag definitions to the haskell codebase (Issue 641)
186 - Fix RAPI network tag handling
187 - Add the network tags to the tags searched by gnt-cluster search-tags
188 - Fix caching bug preventing jobs from being cancelled
189 - Start-master/stop-master was always failing if ConfD was disabled. (Issue 685)
195 *(Released Fri, 13 Dec 2013)*
197 - use custom KVM path if set for version checking
198 - SingleNotifyPipeCondition: don't share pollers
200 Inherited from the 2.8 branch:
202 - Fixed Luxi daemon socket permissions after master-failover
203 - Improve IP version detection code directly checking for colons rather than
204 passing the family from the cluster object
205 - Fix NODE/NODE_RES locking in LUInstanceCreate by not acquiring NODE_RES locks
206 opportunistically anymore (Issue 622)
207 - Allow link local IPv6 gateways (Issue 624)
208 - Fix error printing (Issue 616)
209 - Fix a bug in InstanceSetParams concerning names: in case no name is passed in
210 disk modifications, keep the old one. If name=none then set disk name to
212 - Update build_chroot script to work with the latest hackage packages
213 - Add a packet number limit to "fping" in master-ip-setup (Issue 630)
214 - Fix evacuation out of drained node (Issue 615)
215 - Add default file_driver if missing (Issue 571)
216 - Fix job error message after unclean master shutdown (Issue 618)
217 - Lock group(s) when creating instances (Issue 621)
218 - SetDiskID() before accepting an instance (Issue 633)
219 - Allow the ext template disks to receive arbitrary parameters, both at creation
220 time and while being modified
221 - Xen handle domain shutdown (future proofing cherry-pick)
222 - Refactor reading live data in htools (future proofing cherry-pick)
228 *(Released Wed, 13 Nov 2013)*
230 - fix bug, that kept nodes offline when readding
231 - when verifying DRBD versions, ignore unavailable nodes
232 - fix bug that made the console unavailable on kvm in split-user
234 - DRBD: ensure peers are UpToDate for dual-primary (inherited 2.8.2)
240 *(Released Tue, 5 Nov 2013)*
242 Incompatible/important changes
243 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
245 - hroller now also plans for capacity to move non-redundant instances off
246 any node to be rebooted; the old behavior of completely ignoring any
247 non-redundant instances can be restored by adding the --ignore-non-redundant
249 - The cluster option '--no-lvm-storage' was removed in favor of the new option
250 '--enabled-disk-templates'.
251 - On instance creation, disk templates no longer need to be specified
252 with '-t'. The default disk template will be taken from the list of
253 enabled disk templates.
254 - The monitoring daemon is now running as root, in order to be able to collect
255 information only available to root (such as the state of Xen instances).
256 - The ConfD client is now IPv6 compatible.
257 - File and shared file storage is no longer dis/enabled at configure time,
258 but using the option '--enabled-disk-templates' at cluster initialization and
260 - The default directories for file and shared file storage are not anymore
261 specified at configure time, but taken from the cluster's configuration.
262 They can be set at cluster initialization and modification with
263 '--file-storage-dir' and '--shared-file-storage-dir'.
264 - Cluster verification now includes stricter checks regarding the
265 default file and shared file storage directories. It now checks that
266 the directories are explicitely allowed in the 'file-storage-paths' file and
267 that the directories exist on all nodes.
268 - The list of allowed disk templates in the instance policy and the list
269 of cluster-wide enabled disk templates is now checked for consistency
270 on cluster or group modification. On cluster initialization, the ipolicy
271 disk templates are ensured to be a subset of the cluster-wide enabled
277 - DRBD 8.4 support. Depending on the installed DRBD version, Ganeti now uses
278 the correct command syntax. It is possible to use different DRBD versions
279 on different nodes as long as they are compatible to each other. This
280 enables rolling upgrades of DRBD with no downtime. As permanent operation
281 of different DRBD versions within a node group is discouraged,
282 ``gnt-cluster verify`` will emit a warning if it detects such a situation.
283 - New "inst-status-xen" data collector for the monitoring daemon, providing
284 information about the state of the xen instances on the nodes.
285 - New "lv" data collector for the monitoring daemon, collecting data about the
286 logical volumes on the nodes, and pairing them with the name of the instances
288 - New "diskstats" data collector, collecting the data from /proc/diskstats and
289 presenting them over the monitoring daemon interface.
290 - The ConfD client is now IPv6 compatible.
294 The following new dependencies have been added.
298 - ``python-mock`` (http://www.voidspace.org.uk/python/mock/) is now a required
299 for the unit tests (and only used for testing).
303 - ``hslogger`` (http://software.complete.org/hslogger) is now always
304 required, even if confd is not enabled.
309 - Correctly start/stop luxid during gnt-cluster master-failover (inherited
311 - Improved error messsages (inherited from stable-2.8)
317 *(Released Tue, 15 Oct 2013)*
319 The third release candidate in the 2.9 series. Since 2.9.0 rc2:
321 - in implicit configuration upgrade, match ipolicy with enabled disk templates
322 - improved harep documentation (inherited from stable-2.8)
328 *(Released Wed, 9 Oct 2013)*
330 The second release candidate in the 2.9 series. Since 2.9.0 rc1:
332 - Fix bug in cfgupgrade that led to failure when upgrading from 2.8 with
333 at least one DRBD instance.
334 - Fix bug in cfgupgrade that led to an invalid 2.8 configuration after
341 *(Released Tue, 1 Oct 2013)*
343 The first release candidate in the 2.9 series. Since 2.9.0 beta1:
346 - update of the documentation, in particular installation instructions
347 - merging of LD_* constants into DT_* constants
348 - python style changes to be compatible with newer versions of pylint
354 *(Released Thu, 29 Aug 2013)*
356 This was the first beta release of the 2.9 series. All important changes
357 are listed in the latest 2.9 entry.
363 *(Released Thu, 23 Jan 2014)*
365 - Change the list separator for the usb_devices parameter from comma to space.
366 Commas could not work because they are already the hypervisor option
367 separator (Issue 649)
368 - Add support for blktap2 file-driver (Issue 638)
369 - Add network tag definitions to the haskell codebase (Issue 641)
370 - Fix RAPI network tag handling
371 - Add the network tags to the tags searched by gnt-cluster search-tags
372 - Fix caching bug preventing jobs from being cancelled
373 - Start-master/stop-master was always failing if ConfD was disabled. (Issue 685)
379 *(Released Thu, 12 Dec 2013)*
381 - Fixed Luxi daemon socket permissions after master-failover
382 - Improve IP version detection code directly checking for colons rather than
383 passing the family from the cluster object
384 - Fix NODE/NODE_RES locking in LUInstanceCreate by not acquiring NODE_RES locks
385 opportunistically anymore (Issue 622)
386 - Allow link local IPv6 gateways (Issue 624)
387 - Fix error printing (Issue 616)
388 - Fix a bug in InstanceSetParams concerning names: in case no name is passed in
389 disk modifications, keep the old one. If name=none then set disk name to
391 - Update build_chroot script to work with the latest hackage packages
392 - Add a packet number limit to "fping" in master-ip-setup (Issue 630)
393 - Fix evacuation out of drained node (Issue 615)
394 - Add default file_driver if missing (Issue 571)
395 - Fix job error message after unclean master shutdown (Issue 618)
396 - Lock group(s) when creating instances (Issue 621)
397 - SetDiskID() before accepting an instance (Issue 633)
398 - Allow the ext template disks to receive arbitrary parameters, both at creation
399 time and while being modified
400 - Xen handle domain shutdown (future proofing cherry-pick)
401 - Refactor reading live data in htools (future proofing cherry-pick)
407 *(Released Thu, 07 Nov 2013)*
409 - DRBD: ensure peers are UpToDate for dual-primary
410 - Improve error message for replace-disks
411 - More dependency checks at configure time
412 - Placate warnings on ganeti.outils_unittest.py
418 *(Released Thu, 17 Oct 2013)*
420 - Correctly start/stop luxid during gnt-cluster master-failover
421 - Don't attempt IPv6 ssh in case of IPv4 cluster (Issue 595)
422 - Fix path for the job queue serial file
423 - Improved harep man page
424 - Minor documentation improvements
430 *(Released Mon, 30 Sep 2013)*
432 Incompatible/important changes
433 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
435 - Instance policy can contain multiple instance specs, as described in
436 the “Constrained instance sizes” section of :doc:`Partitioned Ganeti
437 <design-partitioned>`. As a consequence, it's not possible to partially change
438 or override instance specs. Bounding specs (min and max) can be specified as a
439 whole using the new option ``--ipolicy-bounds-specs``, while standard
440 specs use the new option ``--ipolicy-std-specs``.
441 - The output of the info command of gnt-cluster, gnt-group, gnt-node,
442 gnt-instance is a valid YAML object.
443 - hail now honors network restrictions when allocating nodes. This led to an
444 update of the IAllocator protocol. See the IAllocator documentation for
446 - confd now only answers static configuration request over the network. luxid
447 was extracted, listens on the local LUXI socket and responds to live queries.
448 This allows finer grained permissions if using separate users.
453 - The :doc:`Remote API <rapi>` daemon now supports a command line flag
454 to always require authentication, ``--require-authentication``. It can
455 be specified in ``$sysconfdir/default/ganeti``.
456 - A new cluster attribute 'enabled_disk_templates' is introduced. It will
457 be used to manage the disk templates to be used by instances in the cluster.
458 Initially, it will be set to a list that includes plain, drbd, if they were
459 enabled by specifying a volume group name, and file and sharedfile, if those
460 were enabled at configure time. Additionally, it will include all disk
461 templates that are currently used by instances. The order of disk templates
462 will be based on Ganeti's history of supporting them. In the future, the
463 first entry of the list will be used as a default disk template on instance
465 - ``cfgupgrade`` now supports a ``--downgrade`` option to bring the
466 configuration back to the previous stable version.
467 - Disk templates in group ipolicy can be restored to the default value.
468 - Initial support for diskless instances and virtual clusters in QA.
469 - More QA and unit tests for instance policies.
470 - Every opcode now contains a reason trail (visible through ``gnt-job info``)
471 describing why the opcode itself was executed.
472 - The monitoring daemon is now available. It allows users to query the cluster
473 for obtaining information about the status of the system. The daemon is only
474 responsible for providing the information over the network: the actual data
475 gathering is performed by data collectors (currently, only the DRBD status
476 collector is available).
477 - In order to help developers work on Ganeti, a new script
478 (``devel/build_chroot``) is provided, for building a chroot that contains all
479 the required development libraries and tools for compiling Ganeti on a Debian
481 - A new tool, ``harep``, for performing self-repair and recreation of instances
482 in Ganeti has been added.
483 - Split queries are enabled for tags, network, exports, cluster info, groups,
485 - New command ``show-ispecs-cmd`` for ``gnt-cluster`` and ``gnt-group``.
486 It prints the command line to set the current policies, to ease
488 - Add the ``vnet_hdr`` HV parameter for KVM, to control whether the tap
489 devices for KVM virtio-net interfaces will get created with VNET_HDR
490 (IFF_VNET_HDR) support. If set to false, it disables offloading on the
491 virtio-net interfaces, which prevents host kernel tainting and log
492 flooding, when dealing with broken or malicious virtio-net drivers.
493 It's set to true by default.
494 - Instance failover now supports a ``--cleanup`` parameter for fixing previous
496 - Support 'viridian' parameter in Xen HVM
497 - Support DSA SSH keys in bootstrap
498 - To simplify the work of packaging frameworks that want to add the needed users
499 and groups in a split-user setup themselves, at build time three files in
500 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
501 the groups to be generated, the ``users`` file contains, one per line, the
502 users to be generated, optionally followed by their primary group, where
503 important. The ``groupmemberships`` file contains, one per line, additional
504 user-group membership relations that need to be established. The syntax of
505 these files will remain stable in all future versions.
510 The following new dependencies have been added:
513 - The ``curl`` library is not optional anymore for compiling the Haskell code.
514 - ``snap-server`` library (if monitoring is enabled).
517 - The minimum Python version needed to run Ganeti is now 2.6.
518 - ``yaml`` library (only for running the QA).
522 - Perform proper cleanup on termination of Haskell daemons
523 - Fix corner-case in handling of remaining retry time
529 *(Released Tue, 17 Sep 2013)*
531 - To simplify the work of packaging frameworks that want to add the needed users
532 and groups in a split-user setup themselves, at build time three files in
533 ``doc/users`` will be generated. The ``groups`` files contains, one per line,
534 the groups to be generated, the ``users`` file contains, one per line, the
535 users to be generated, optionally followed by their primary group, where
536 important. The ``groupmemberships`` file contains, one per line, additional
537 user-group membership relations that need to be established. The syntax of
538 these files will remain stable in all future versions.
539 - Add a default to file-driver when unspecified over RAPI (Issue 571)
540 - Mark the DSA host pubkey as optional, and remove it during config downgrade
542 - Some documentation fixes
548 *(Released Tue, 27 Aug 2013)*
550 The second release candidate of the 2.8 series. Since 2.8.0. rc1:
552 - Support 'viridian' parameter in Xen HVM (Issue 233)
553 - Include VCS version in ``gnt-cluster version``
554 - Support DSA SSH keys in bootstrap (Issue 338)
555 - Fix batch creation of instances
556 - Use FQDN to check master node status (Issue 551)
557 - Make the DRBD collector more failure-resilient
563 *(Released Fri, 2 Aug 2013)*
565 The first release candidate of the 2.8 series. Since 2.8.0 beta1:
567 - Fix upgrading/downgrading from 2.7
568 - Increase maximum RAPI message size
569 - Documentation updates
570 - Split ``confd`` between ``luxid`` and ``confd``
571 - Merge 2.7 series up to the 2.7.1 release
572 - Allow the ``modify_etc_hosts`` option to be changed
573 - Add better debugging for ``luxid`` queries
574 - Expose bulk parameter for GetJobs in RAPI client
575 - Expose missing ``network`` fields in RAPI
576 - Add some ``cluster verify`` tests
577 - Some unittest fixes
578 - Fix a malfunction in ``hspace``'s tiered allocation
579 - Fix query compatibility between haskell and python implementations
580 - Add the ``vnet_hdr`` HV parameter for KVM
581 - Add ``--cleanup`` to instance failover
582 - Change the connected groups format in ``gnt-network info`` output; it
583 was previously displayed as a raw list by mistake. (Merged from 2.7)
589 *(Released Mon, 24 Jun 2013)*
591 This was the first beta release of the 2.8 series. All important changes
592 are listed in the latest 2.8 entry.
598 *(Released Thu, 26 Sep 2013)*
600 - Change the connected groups format in ``gnt-network info`` output; it
601 was previously displayed as a raw list by mistake
602 - Check disk template in right dict when copying
603 - Support multi-instance allocs without iallocator
604 - Fix some errors in the documentation
605 - Fix formatting of tuple in an error message
611 *(Released Thu, 25 Jul 2013)*
613 - Add logrotate functionality in daemon-util
614 - Add logrotate example file
615 - Add missing fields to network queries over rapi
616 - Fix network object timestamps
617 - Add support for querying network timestamps
618 - Fix a typo in the example crontab
619 - Fix a documentation typo
625 *(Released Thu, 04 Jul 2013)*
627 Incompatible/important changes
628 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
630 - Instance policies for disk size were documented to be on a per-disk
631 basis, but hail applied them to the sum of all disks. This has been
633 - ``hbal`` will now exit with status 0 if, during job execution over
634 LUXI, early exit has been requested and all jobs are successful;
635 before, exit status 1 was used, which cannot be differentiated from
637 - Compatibility with newer versions of rbd has been fixed
638 - ``gnt-instance batch-create`` has been changed to use the bulk create
639 opcode from Ganeti. This lead to incompatible changes in the format of
640 the JSON file. It's now not a custom dict anymore but a dict
641 compatible with the ``OpInstanceCreate`` opcode.
642 - Parent directories for file storage need to be listed in
643 ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
644 write the file automatically based on old configuration values, but it
645 can not distribute it across all nodes and the file contents should be
646 verified. Use ``gnt-cluster copyfile
647 $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
648 upgraded. The reason for requiring this list of paths now is that
649 before it would have been possible to inject new paths via RPC,
650 allowing files to be created in arbitrary locations. The RPC protocol
651 is protected using SSL/X.509 certificates, but as a design principle
652 Ganeti does not permit arbitrary paths to be passed.
653 - The parsing of the variants file for OSes (see
654 :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
655 lines and comment lines (starting with ``#``) are ignored for better
657 - The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
658 longer available. ``gnt-node add`` now invokes a new tool on the
659 destination node, named ``prepare-node-join``, to configure the SSH
660 daemon. Paramiko is no longer necessary to configure nodes' SSH
661 daemons via ``gnt-node add``.
662 - Draining (``gnt-cluster queue drain``) and un-draining the job queue
663 (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
664 the flag is not reset after a master failover.
665 - Python 2.4 has *not* been tested with this release. Using 2.6 or above
666 is recommended. 2.6 will be mandatory from the 2.8 series.
672 - New network management functionality to support automatic allocation
673 of IP addresses and managing of network parameters. See
674 :manpage:`gnt-network(8)` for more details.
675 - New external storage backend, to allow managing arbitrary storage
676 systems external to the cluster. See
677 :manpage:`ganeti-extstorage-interface(7)`.
678 - New ``exclusive-storage`` node parameter added, restricted to
679 nodegroup level. When it's set to true, physical disks are assigned in
680 an exclusive fashion to instances, as documented in :doc:`Partitioned
681 Ganeti <design-partitioned>`. Currently, only instances using the
682 ``plain`` disk template are supported.
683 - The KVM hypervisor has been updated with many new hypervisor
684 parameters, including a generic one for passing arbitrary command line
685 values. See a complete list in :manpage:`gnt-instance(8)`. It is now
686 compatible up to qemu 1.4.
687 - A new tool, called ``mon-collector``, is the stand-alone executor of
688 the data collectors for a monitoring system. As of this version, it
689 just includes the DRBD data collector, that can be executed by calling
690 ``mon-collector`` using the ``drbd`` parameter. See
691 :manpage:`mon-collector(7)`.
692 - A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
693 for RAPI users. It allows granting permissions to query for
694 information to a specific user without giving
695 :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
696 - A new tool named ``node-cleanup`` has been added. It cleans remains of
697 a cluster from a machine by stopping all daemons, removing
698 certificates and ssconf files. Unless the ``--no-backup`` option is
699 given, copies of the certificates are made.
700 - Instance creations now support the use of opportunistic locking,
701 potentially speeding up the (parallel) creation of multiple instances.
702 This feature is currently only available via the :doc:`RAPI
703 <rapi>` interface and when an instance allocator is used. If the
704 ``opportunistic_locking`` parameter is set the opcode will try to
705 acquire as many locks as possible, but will not wait for any locks
706 held by other opcodes. If not enough resources can be found to
707 allocate the instance, the temporary error code
708 :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
709 retried thereafter, with or without opportunistic locking.
710 - New experimental linux-ha resource scripts.
711 - Restricted-commands support: ganeti can now be asked (via command line
712 or rapi) to perform commands on a node. These are passed via ganeti
713 RPC rather than ssh. This functionality is restricted to commands
714 specified on the ``$sysconfdir/ganeti/restricted-commands`` for security
715 reasons. The file is not copied automatically.
721 - Diskless instances are now externally mirrored (Issue 237). This for
722 now has only been tested in conjunction with explicit target nodes for
724 - Queries not needing locks or RPC access to the node can now be
725 performed by the confd daemon, making them independent from jobs, and
726 thus faster to execute. This is selectable at configure time.
727 - The functionality for allocating multiple instances at once has been
728 overhauled and is now also available through :doc:`RAPI <rapi>`.
730 There are no significant changes from version 2.7.0~rc3.
736 *(Released Tue, 25 Jun 2013)*
738 - Fix permissions on the confd query socket (Issue 477)
739 - Fix permissions on the job archive dir (Issue 498)
740 - Fix handling of an internal exception in replace-disks (Issue 472)
741 - Fix gnt-node info handling of shortened names (Issue 497)
742 - Fix gnt-instance grow-disk when wiping is enabled
743 - Documentation improvements, and support for newer pandoc
744 - Fix hspace honoring ipolicy for disks (Issue 484)
745 - Improve handling of the ``kvm_extra`` HV parameter
751 *(Released Fri, 24 May 2013)*
753 - ``devel/upload`` now works when ``/var/run`` on the target nodes is a
755 - Disks added through ``gnt-instance modify`` or created through
756 ``gnt-instance recreate-disks`` are wiped, if the
757 ``prealloc_wipe_disks`` flag is set.
758 - If wiping newly created disks fails, the disks are removed. Also,
759 partial failures in creating disks through ``gnt-instance modify``
760 triggers a cleanup of the partially-created disks.
761 - Removing the master IP address doesn't fail if the address has been
763 - Fix ownership of the OS log dir
764 - Workaround missing SO_PEERCRED constant (Issue 191)
770 *(Released Fri, 3 May 2013)*
772 This was the first release candidate of the 2.7 series. Since beta3:
774 - Fix kvm compatibility with qemu 1.4 (Issue 389)
775 - Documentation updates (admin guide, upgrade notes, install
776 instructions) (Issue 372)
777 - Fix gnt-group list nodes and instances count (Issue 436)
778 - Fix compilation without non-mandatory libraries (Issue 441)
779 - Fix xen-hvm hypervisor forcing nics to type 'ioemu' (Issue 247)
780 - Make confd logging more verbose at INFO level (Issue 435)
781 - Improve "networks" documentation in :manpage:`gnt-instance(8)`
782 - Fix failure path for instance storage type conversion (Issue 229)
783 - Update htools text backend documentation
784 - Improve the renew-crypto section of :manpage:`gnt-cluster(8)`
785 - Disable inter-cluster instance move for file-based instances, because
786 it is dependant on instance export, which is not supported for
787 file-based instances. (Issue 414)
788 - Fix gnt-job crashes on non-ascii characters (Issue 427)
789 - Fix volume group checks on non-vm-capable nodes (Issue 432)
795 *(Released Mon, 22 Apr 2013)*
797 This was the third beta release of the 2.7 series. Since beta2:
799 - Fix hail to verify disk instance policies on a per-disk basis (Issue 418).
800 - Fix data loss on wrong usage of ``gnt-instance move``
801 - Properly export errors in confd-based job queries
802 - Add ``users-setup`` tool
803 - Fix iallocator protocol to report 0 as a disk size for diskless
804 instances. This avoids hail breaking when a diskless instance is
806 - Fix job queue directory permission problem that made confd job queries
807 fail. This requires running an ``ensure-dirs --full-run`` on upgrade
808 for access to archived jobs (Issue 406).
809 - Limit the sizes of networks supported by ``gnt-network`` to something
810 between a ``/16`` and a ``/30`` to prevent memory bloat and crashes.
811 - Fix bugs in instance disk template conversion
812 - Fix GHC 7 compatibility
813 - Fix ``burnin`` install path (Issue 426).
814 - Allow very small disk grows (Issue 347).
815 - Fix a ``ganeti-noded`` memory bloat introduced in 2.5, by making sure
816 that noded doesn't import masterd code (Issue 419).
817 - Make sure the default metavg at cluster init is the same as the vg, if
818 unspecified (Issue 358).
819 - Fix cleanup of partially created disks (part of Issue 416)
825 *(Released Tue, 2 Apr 2013)*
827 This was the second beta release of the 2.7 series. Since beta1:
829 - Networks no longer have a "type" slot, since this information was
830 unused in Ganeti: instead of it tags should be used.
831 - The rapi client now has a ``target_node`` option to MigrateInstance.
832 - Fix early exit return code for hbal (Issue 386).
833 - Fix ``gnt-instance migrate/failover -n`` (Issue 396).
834 - Fix ``rbd showmapped`` output parsing (Issue 312).
835 - Networks are now referenced indexed by UUID, rather than name. This
836 will require running cfgupgrade, from 2.7.0beta1, if networks are in
838 - The OS environment now includes network information.
839 - Deleting of a network is now disallowed if any instance nic is using
840 it, to prevent dangling references.
841 - External storage is now documented in man pages.
842 - The exclusive_storage flag can now only be set at nodegroup level.
843 - Hbal can now submit an explicit priority with its jobs.
844 - Many network related locking fixes.
845 - Bump up the required pylint version to 0.25.1.
846 - Fix the ``no_remember`` option in RAPI client.
847 - Many ipolicy related tests, qa, and fixes.
848 - Many documentation improvements and fixes.
849 - Fix building with ``--disable-file-storage``.
850 - Fix ``-q`` option in htools, which was broken if passed more than
852 - Some haskell/python interaction improvements and fixes.
853 - Fix iallocator in case of missing LVM storage.
854 - Fix confd config load in case of ``--no-lvm-storage``.
855 - The confd/query functionality is now mentioned in the security
862 *(Released Wed, 6 Feb 2013)*
864 This was the first beta release of the 2.7 series. All important changes
865 are listed in the latest 2.7 entry.
871 *(Released Fri, 21 Dec 2012)*
873 Important behaviour change: hbal won't rebalance anymore instances which
874 have the ``auto_balance`` attribute set to false. This was the intention
875 all along, but until now it only skipped those from the N+1 memory
876 reservation (DRBD-specific).
878 A significant number of bug fixes in this release:
880 - Fixed disk adoption interaction with ipolicy checks.
881 - Fixed networking issues when instances are started, stopped or
882 migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
883 - Fixed the warning in cluster verify for shared storage instances not
885 - Fixed removal of storage directory on shared file storage (issue 262).
886 - Fixed validation of LVM volume group name in OpClusterSetParams
887 (``gnt-cluster modify``) (issue 285).
888 - Fixed runtime memory increases (``gnt-instance modify -m``).
889 - Fixed live migration under Xen's ``xl`` mode.
890 - Fixed ``gnt-instance console`` with ``xl``.
891 - Fixed building with newer Haskell compiler/libraries.
892 - Fixed PID file writing in Haskell daemons (confd); this prevents
893 restart issues if confd was launched manually (outside of
894 ``daemon-util``) while another copy of it was running
895 - Fixed a type error when doing live migrations with KVM (issue 297) and
896 the error messages for failing migrations have been improved.
897 - Fixed opcode validation for the out-of-band commands (``gnt-node
899 - Fixed a type error when unsetting OS hypervisor parameters (issue
900 311); now it's possible to unset all OS-specific hypervisor
902 - Fixed the ``dry-run`` mode for many operations: verification of
903 results was over-zealous but didn't take into account the ``dry-run``
904 operation, resulting in "wrong" failures.
905 - Fixed bash completion in ``gnt-job list`` when the job queue has
906 hundreds of entries; especially with older ``bash`` versions, this
907 results in significant CPU usage.
909 And lastly, a few other improvements have been made:
911 - Added option to force master-failover without voting (issue 282).
912 - Clarified error message on lock conflict (issue 287).
913 - Logging of newly submitted jobs has been improved (issue 290).
914 - Hostname checks have been made uniform between instance rename and
916 - The ``--submit`` option is now supported by ``gnt-debug delay``.
917 - Shutting down the master daemon by sending SIGTERM now stops it from
918 processing jobs waiting for locks; instead, those jobs will be started
919 once again after the master daemon is started the next time (issue
921 - Support for Xen's ``xl`` program has been improved (besides the fixes
923 - Reduced logging noise in the Haskell confd daemon (only show one log
924 entry for each config reload, instead of two).
925 - Several man page updates and typo fixes.
931 *(Released Fri, 12 Oct 2012)*
933 A small bugfix release. Among the bugs fixed:
935 - Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
936 made the command unusable.
937 - Commands that issue many jobs don't fail anymore just because some jobs
938 take so long that other jobs are archived.
939 - Failures during ``gnt-instance reinstall`` are reflected by the exit
941 - Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
943 - When ``always_failover`` is set, ``--allow-failover`` is not required
944 in migrate commands anymore.
945 - ``bash_completion`` works even if extglob is disabled.
946 - Fixed bug with locks that made failover for RDB-based instances fail.
947 - Fixed bug in non-mirrored instance allocation that made Ganeti choose
948 a random node instead of one based on the allocator metric.
949 - Support for newer versions of pylint and pep8.
950 - Hail doesn't fail anymore when trying to add an instance of type
951 ``file``, ``sharedfile`` or ``rbd``.
952 - Added new Makefile target to rebuild the whole distribution, so that
953 all files are included.
959 *(Released Fri, 27 Jul 2012)*
962 .. attention:: The ``LUXI`` protocol has been made more consistent
963 regarding its handling of command arguments. This, however, leads to
964 incompatibility issues with previous versions. Please ensure that you
965 restart Ganeti daemons soon after the upgrade, otherwise most
966 ``LUXI`` calls (job submission, setting/resetting the drain flag,
967 pausing/resuming the watcher, cancelling and archiving jobs, querying
968 the cluster configuration) will fail.
977 The current ``admin_up`` field, which used to denote whether an instance
978 should be running or not, has been removed. Instead, ``admin_state`` is
979 introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
981 The rational behind this is that an instance being “down” can have
984 - it could be down during a reboot
985 - it could be temporarily be down for a reinstall
986 - or it could be down because it is deprecated and kept just for its
989 The previous Boolean state was making it difficult to do capacity
990 calculations: should Ganeti reserve memory for a down instance? Now, the
991 tri-state field makes it clear:
993 - in ``up`` and ``down`` state, all resources are reserved for the
994 instance, and it can be at any time brought up if it is down
995 - in ``offline`` state, only disk space is reserved for it, but not
998 The field can have an extra use: since the transition between ``up`` and
999 ``down`` and vice-versus is done via ``gnt-instance start/stop``, but
1000 transition between ``offline`` and ``down`` is done via ``gnt-instance
1001 modify``, it is possible to given different rights to users. For
1002 example, owners of an instance could be allowed to start/stop it, but
1003 not transition it out of the offline state.
1005 Instance policies and specs
1006 +++++++++++++++++++++++++++
1008 In previous Ganeti versions, an instance creation request was not
1009 limited on the minimum size and on the maximum size just by the cluster
1010 resources. As such, any policy could be implemented only in third-party
1011 clients (RAPI clients, or shell wrappers over ``gnt-*``
1012 tools). Furthermore, calculating cluster capacity via ``hspace`` again
1013 required external input with regards to instance sizes.
1015 In order to improve these workflows and to allow for example better
1016 per-node group differentiation, we introduced instance specs, which
1019 - minimum instance disk size, disk count, memory size, cpu count
1020 - maximum values for the above metrics
1021 - and “standard” values (used in ``hspace`` to calculate the standard
1024 The minimum/maximum values can be also customised at node-group level,
1025 for example allowing more powerful hardware to support bigger instance
1028 Beside the instance specs, there are a few other settings belonging to
1029 the instance policy framework. It is possible now to customise, per
1030 cluster and node-group:
1032 - the list of allowed disk templates
1033 - the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
1034 - the maximum ratio of instance to spindles (see below for more
1035 information) for local storage
1037 All these together should allow all tools that talk to Ganeti to know
1038 what are the ranges of allowed values for instances and the
1039 over-subscription that is allowed.
1041 For the VCPU/PCPU ratio, we already have the VCPU configuration from the
1042 instance configuration, and the physical CPU configuration from the
1043 node. For the spindle ratios however, we didn't track before these
1044 values, so new parameters have been added:
1046 - a new node parameter ``spindle_count``, defaults to 1, customisable at
1047 node group or node level
1048 - at new backend parameter (for instances), ``spindle_use`` defaults to 1
1050 Note that spindles in this context doesn't need to mean actual
1051 mechanical hard-drives; it's just a relative number for both the node
1052 I/O capacity and instance I/O consumption.
1054 Instance migration behaviour
1055 ++++++++++++++++++++++++++++
1057 While live-migration is in general desirable over failover, it is
1058 possible that for some workloads it is actually worse, due to the
1059 variable time of the “suspend” phase during live migration.
1061 To allow the tools to work consistently over such instances (without
1062 having to hard-code instance names), a new backend parameter
1063 ``always_failover`` has been added to control the migration/failover
1064 behaviour. When set to True, all migration requests for an instance will
1065 instead fall-back to failover.
1067 Instance memory ballooning
1068 ++++++++++++++++++++++++++
1070 Initial support for memory ballooning has been added. The memory for an
1071 instance is no longer fixed (backend parameter ``memory``), but instead
1072 can vary between minimum and maximum values (backend parameters
1073 ``minmem`` and ``maxmem``). Currently we only change an instance's
1076 - live migrating or failing over and instance and the target node
1077 doesn't have enough memory
1078 - user requests changing the memory via ``gnt-instance modify
1081 Instance CPU pinning
1082 ++++++++++++++++++++
1084 In order to control the use of specific CPUs by instance, support for
1085 controlling CPU pinning has been added for the Xen, HVM and LXC
1086 hypervisors. This is controlled by a new hypervisor parameter
1087 ``cpu_mask``; details about possible values for this are in the
1088 :manpage:`gnt-instance(8)`. Note that use of the most specific (precise
1089 VCPU-to-CPU mapping) form will work well only when all nodes in your
1090 cluster have the same amount of CPUs.
1095 Another area in which Ganeti was not customisable were the parameters
1096 used for storage configuration, e.g. how many stripes to use for LVM,
1097 DRBD resync configuration, etc.
1099 To improve this area, we've added disks parameters, which are
1100 customisable at cluster and node group level, and which allow to
1101 specify various parameters for disks (DRBD has the most parameters
1102 currently), for example:
1104 - DRBD resync algorithm and parameters (e.g. speed)
1105 - the default VG for meta-data volumes for DRBD
1106 - number of stripes for LVM (plain disk template)
1109 These parameters can be modified via ``gnt-cluster modify -D …`` and
1110 ``gnt-group modify -D …``, and are used at either instance creation (in
1111 case of LVM stripes, for example) or at disk “activation” time
1112 (e.g. resync speed).
1114 Rados block device support
1115 ++++++++++++++++++++++++++
1117 A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
1118 denoted by the ``rbd`` disk template type. This is considered
1119 experimental, feedback is welcome. For details on configuring it, see
1120 the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
1125 The existing master IP functionality works well only in simple setups (a
1126 single network shared by all nodes); however, if nodes belong to
1127 different networks, then the ``/32`` setup and lack of routing
1128 information is not enough.
1130 To allow the master IP to function well in more complex cases, the
1131 system was reworked as follows:
1133 - a master IP netmask setting has been added
1134 - the master IP activation/turn-down code was moved from the node daemon
1135 to a separate script
1136 - whether to run the Ganeti-supplied master IP script or a user-supplied
1137 on is a ``gnt-cluster init`` setting
1139 Details about the location of the standard and custom setup scripts are
1140 in the man page :manpage:`gnt-cluster(8)`; for information about the
1141 setup script protocol, look at the Ganeti-supplied script.
1146 The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
1149 It is now possible to use TLS-protected connections, and when renewing
1150 or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
1151 it is now possible to specify spice or spice CA certificates. Also, it
1152 is possible to configure a password for SPICE sessions via the
1153 hypervisor parameter ``spice_password_file``.
1155 There are also new parameters to control the compression and streaming
1156 options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
1157 etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
1158 for the spice parameters.
1160 Lastly, it is now possible to see the SPICE connection information via
1161 ``gnt-instance console``.
1166 A new tool (``tools/ovfconverter``) has been added that supports
1167 conversion between Ganeti and the `Open Virtualization Format
1168 <http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
1171 This relies on the ``qemu-img`` tool to convert the disk formats, so the
1172 actual compatibility with other virtualization solutions depends on it.
1174 Confd daemon changes
1175 ++++++++++++++++++++
1177 The configuration query daemon (``ganeti-confd``) is now optional, and
1178 has been rewritten in Haskell; whether to use the daemon at all, use the
1179 Python (default) or the Haskell version is selectable at configure time
1180 via the ``--enable-confd`` parameter, which can take one of the
1181 ``haskell``, ``python`` or ``no`` values. If not used, disabling the
1182 daemon will result in a smaller footprint; for larger systems, we
1183 welcome feedback on the Haskell version which might become the default
1186 If you want to use ``gnt-node list-drbd`` you need to have the Haskell
1187 daemon running. The Python version doesn't implement the new call.
1190 User interface changes
1191 ~~~~~~~~~~~~~~~~~~~~~~
1193 We have replaced the ``--disks`` option of ``gnt-instance
1194 replace-disks`` with a more flexible ``--disk`` option, which allows
1195 adding and removing disks at arbitrary indices (Issue 188). Furthermore,
1196 disk size and mode can be changed upon recreation (via ``gnt-instance
1197 recreate-disks``, which accepts the same ``--disk`` option).
1199 As many people are used to a ``show`` command, we have added that as an
1200 alias to ``info`` on all ``gnt-*`` commands.
1202 The ``gnt-instance grow-disk`` command has a new mode in which it can
1203 accept the target size of the disk, instead of the delta; this can be
1204 more safe since two runs in absolute mode will be idempotent, and
1205 sometimes it's also easier to specify the desired size directly.
1207 Also the handling of instances with regard to offline secondaries has
1208 been improved. Instance operations should not fail because one of it's
1209 secondary nodes is offline, even though it's safe to proceed.
1211 A new command ``list-drbd`` has been added to the ``gnt-node`` script to
1212 support debugging of DRBD issues on nodes. It provides a mapping of DRBD
1213 minors to instance name.
1218 RAPI coverage has improved, with (for example) new resources for
1219 recreate-disks, node power-cycle, etc.
1224 There is partial support for ``xl`` in the Xen hypervisor; feedback is
1227 Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
1228 whether to still support Python 2.4 or move to Python 2.6 as minimum
1231 Support for Fedora has been slightly improved; the provided example
1232 init.d script should work better on it and the INSTALL file should
1233 document the needed dependencies.
1238 The deprecated ``QueryLocks`` LUXI request has been removed. Use
1239 ``Query(what=QR_LOCK, ...)`` instead.
1241 The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
1242 :pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
1243 :pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
1244 :pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
1245 future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
1247 RAPI client: ``CertificateError`` now derives from
1248 ``GanetiApiError``. This should make it more easy to handle Ganeti
1251 Deprecation warnings due to PyCrypto/paramiko import in
1252 ``tools/setup-ssh`` have been silenced, as usually they are safe; please
1253 make sure to run an up-to-date paramiko version, if you use this tool.
1255 The QA scripts now depend on Python 2.5 or above (the main code base
1256 still works with Python 2.4).
1258 The configuration file (``config.data``) is now written without
1259 indentation for performance reasons; if you want to edit it, it can be
1260 re-formatted via ``tools/fmtjson``.
1262 A number of bugs has been fixed in the cluster merge tool.
1264 ``x509`` certification verification (used in import-export) has been
1265 changed to allow the same clock skew as permitted by the cluster
1266 verification. This will remove some rare but hard to diagnose errors in
1273 *(Released Thu, 19 Jul 2012)*
1275 Very few changes from rc4 to the final release, only bugfixes:
1277 - integrated fixes from release 2.5.2 (fix general boot flag for KVM
1278 instance, fix CDROM booting for KVM instances)
1279 - fixed node group modification of node parameters
1280 - fixed issue in LUClusterVerifyGroup with multi-group clusters
1281 - fixed generation of bash completion to ensure a stable ordering
1288 *(Released Fri, 13 Jul 2012)*
1290 Third release candidate for 2.6. The following changes were done from
1293 - Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
1294 - Fixed an inconsistency in the LUXI protocol with the provided
1295 arguments (NOT backwards compatible)
1296 - Fixed a bug with node groups ipolicy where ``min`` was greater than
1297 the cluster ``std`` value
1298 - Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
1299 easier instance debugging on nodes (requires ``hconfd`` to work)
1305 *(Released Tue, 03 Jul 2012)*
1307 Second release candidate for 2.6. The following changes were done from
1310 - Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
1312 - Fixed a RAPI regression on missing beparams/memory
1313 - Fixed redistribution of files on offline nodes
1314 - Added possibility to run activate-disks even though secondaries are
1315 offline. With this change it relaxes also the strictness on some other
1316 commands which use activate disks internally:
1317 * ``gnt-instance start|reboot|rename|backup|export``
1318 - Made it possible to remove safely an instance if its secondaries are
1320 - Made it possible to reinstall even though secondaries are offline
1326 *(Released Mon, 25 Jun 2012)*
1328 First release candidate for 2.6. The following changes were done from
1331 - Fixed bugs with disk parameters and ``rbd`` templates as well as
1333 - Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
1334 behaviour. It supports now the modify operation
1335 - ``hcheck`` implemented to analyze cluster health and possibility of
1336 improving health by rebalance
1337 - ``hbal`` has been improved in dealing with split instances
1343 *(Released Mon, 11 Jun 2012)*
1345 Second beta release of 2.6. The following changes were done from beta2
1348 - Fixed ``daemon-util`` with non-root user models
1349 - Fixed creation of plain instances with ``--no-wait-for-sync``
1350 - Fix wrong iv_names when running ``cfgupgrade``
1351 - Export more information in RAPI group queries
1352 - Fixed bug when changing instance network interfaces
1353 - Extended burnin to do NIC changes
1354 - query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
1355 - Changed default for DRBD barriers
1356 - Fixed DRBD error reporting for syncer rate
1357 - Verify the options on disk parameters
1359 And of course various fixes to documentation and improved unittests and
1366 *(Released Wed, 23 May 2012)*
1368 First beta release of 2.6. The following changes were done from beta1 to
1371 - integrated patch for distributions without ``start-stop-daemon``
1372 - adapted example init.d script to work on Fedora
1373 - fixed log handling in Haskell daemons
1374 - adapted checks in the watcher for pycurl linked against libnss
1375 - add partial support for ``xl`` instead of ``xm`` for Xen
1376 - fixed a type issue in cluster verification
1377 - fixed ssconf handling in the Haskell code (was breaking confd in IPv6
1380 Plus integrated fixes from the 2.5 branch:
1382 - fixed ``kvm-ifup`` to use ``/bin/bash``
1383 - fixed parallel build failures
1384 - KVM live migration when using a custom keymap
1390 *(Released Tue, 24 Jul 2012)*
1392 A small bugfix release, with no new features:
1394 - fixed bash-isms in kvm-ifup, for compatibility with systems which use a
1395 different default shell (e.g. Debian, Ubuntu)
1396 - fixed KVM startup and live migration with a custom keymap (fixes Issue
1397 243 and Debian bug #650664)
1398 - fixed compatibility with KVM versions that don't support multiple boot
1399 devices (fixes Issue 230 and Debian bug #624256)
1401 Additionally, a few fixes were done to the build system (fixed parallel
1402 build failures) and to the unittests (fixed race condition in test for
1403 FileID functions, and the default enable/disable mode for QA test is now
1410 *(Released Fri, 11 May 2012)*
1412 A small bugfix release.
1414 The main issues solved are on the topic of compatibility with newer LVM
1417 - fixed parsing of ``lv_attr`` field
1418 - adapted to new ``vgreduce --removemissing`` behaviour where sometimes
1419 the ``--force`` flag is needed
1421 Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
1422 to accept kernel 3.x too (was hardcoded to 2.6.*).
1424 A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
1425 of hook results and that also made display of other errors suboptimal
1426 was fixed; the code behaves now like 2.4 and earlier.
1428 Another change in 2.5, the cleanup of the OS scripts environment, is too
1429 aggressive: it removed even the ``PATH`` variable, which requires the OS
1430 scripts to *always* need to export it. Since this is a bit too strict,
1431 we now export a minimal PATH, the same that we export for hooks.
1433 The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
1434 integrated into this release.
1436 Finally, a few other miscellaneous changes were done (no new features,
1437 just small improvements):
1439 - Fix ``gnt-group --help`` display
1440 - Fix hardcoded Xen kernel path
1441 - Fix grow-disk handling of invalid units
1442 - Update synopsis for ``gnt-cluster repair-disk-sizes``
1443 - Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
1449 *(Released Thu, 12 Apr 2012)*
1451 Incompatible/important changes and bugfixes
1452 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1454 - The default of the ``/2/instances/[instance_name]/rename`` RAPI
1455 resource's ``ip_check`` parameter changed from ``True`` to ``False``
1456 to match the underlying LUXI interface.
1457 - The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
1458 body parameters, see :doc:`RAPI documentation <rapi>`. The server does
1459 not maintain backwards-compatibility as the underlying operation
1460 changed in an incompatible way. The RAPI client can talk to old
1461 servers, but it needs to be told so as the return value changed.
1462 - When creating file-based instances via RAPI, the ``file_driver``
1463 parameter no longer defaults to ``loop`` and must be specified.
1464 - The deprecated ``bridge`` NIC parameter is no longer supported. Use
1466 - Support for the undocumented and deprecated RAPI instance creation
1467 request format version 0 has been dropped. Use version 1, supported
1468 since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
1469 - Pyparsing 1.4.6 or above is required, see :doc:`installation
1470 documentation <install>`.
1471 - The "cluster-verify" hooks are now executed per group by the
1472 ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
1473 if you just run ``gnt-cluster verify``, which generates one opcode per
1475 - The environment as passed to the OS scripts is cleared, and thus no
1476 environment variables defined in the node daemon's environment will be
1477 inherited by the scripts.
1478 - The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
1480 - :doc:`New iallocator modes <design-multi-reloc>` have been added to
1481 support operations involving multiple node groups.
1482 - Offline nodes are ignored when failing over an instance.
1483 - Support for KVM version 1.0, which changed the version reporting format
1485 - TCP/IP ports used by DRBD disks are returned to a pool upon instance
1487 - ``Makefile`` is now compatible with Automake 1.11.2
1488 - Includes all bugfixes made in the 2.4 series
1493 - The ganeti-htools project has been merged into the ganeti-core source
1494 tree and will be built as part of Ganeti (see :doc:`install-quick`).
1495 - Implemented support for :doc:`shared storage <design-shared-storage>`.
1496 - Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
1497 GPT-style partition tables (requires `parted
1498 <http://www.gnu.org/s/parted/>`_).
1499 - Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
1500 - Allowed adding tags on instance creation.
1501 - Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
1502 - Allow instances to be started in a paused state, enabling the user to
1503 see the complete console output on boot using the console.
1504 - Added new hypervisor flag to control default reboot behaviour
1505 (``reboot_behavior``).
1506 - Added support for KVM keymaps (hypervisor parameter ``keymap``).
1507 - Improved out-of-band management support:
1509 - Added ``gnt-node health`` command reporting the health status of
1511 - Added ``gnt-node power`` command to manage power status of nodes.
1512 - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
1514 - Instance migration can fall back to failover if instance is not
1516 - Filters can be used when listing nodes, instances, groups and locks;
1517 see :manpage:`ganeti(7)` manpage.
1518 - Added post-execution status as variables to :doc:`hooks <hooks>`
1520 - Instance tags are exported/imported together with the instance.
1521 - When given an explicit job ID, ``gnt-job info`` will work for archived
1523 - Jobs can define dependencies on other jobs (not yet supported via
1524 RAPI or command line, but used by internal commands and usable via
1527 - Lock monitor (``gnt-debug locks``) shows jobs waiting for
1530 - Instance failover is now available as a RAPI resource
1531 (``/2/instances/[instance_name]/failover``).
1532 - ``gnt-instance info`` defaults to static information if primary node
1534 - Opcodes have a new ``comment`` attribute.
1535 - Added basic SPICE support to KVM hypervisor.
1536 - ``tools/ganeti-listrunner`` allows passing of arguments to executable.
1538 Node group improvements
1539 ~~~~~~~~~~~~~~~~~~~~~~~
1541 - ``gnt-cluster verify`` has been modified to check groups separately,
1542 thereby improving performance.
1543 - Node group support has been added to ``gnt-cluster verify-disks``,
1544 which now operates per node group.
1545 - Watcher has been changed to work better with node groups.
1547 - One process and state file per node group.
1548 - Slow watcher in one group doesn't block other group's watcher.
1550 - Added new command, ``gnt-group evacuate``, to move all instances in a
1551 node group to other groups.
1552 - Added ``gnt-instance change-group`` to move an instance to another
1554 - ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
1555 per-group operations.
1556 - Node groups can be tagged.
1557 - Some operations switch from an exclusive to a shared lock as soon as
1559 - Instance's primary and secondary nodes' groups are now available as
1560 query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
1561 and ``snodes.group.uuid``).
1566 - Numerous updates to documentation and manpages.
1568 - :doc:`RAPI <rapi>` documentation now has detailed parameter
1570 - Some opcode/job results are now also documented, see :doc:`RAPI
1573 - A lockset's internal lock is now also visible in lock monitor.
1574 - Log messages from job queue workers now contain information about the
1575 opcode they're processing.
1576 - ``gnt-instance console`` no longer requires the instance lock.
1577 - A short delay when waiting for job changes reduces the number of LUXI
1578 requests significantly.
1579 - DRBD metadata volumes are overwritten with zeros during disk creation.
1580 - Out-of-band commands no longer acquire the cluster lock in exclusive
1582 - ``devel/upload`` now uses correct permissions for directories.
1588 *(Released Fri, 23 Mar 2012)*
1590 This was the sixth release candidate of the 2.5 series.
1596 *(Released Mon, 9 Jan 2012)*
1598 This was the fifth release candidate of the 2.5 series.
1604 *(Released Thu, 27 Oct 2011)*
1606 This was the fourth release candidate of the 2.5 series.
1612 *(Released Wed, 26 Oct 2011)*
1614 This was the third release candidate of the 2.5 series.
1620 *(Released Tue, 18 Oct 2011)*
1622 This was the second release candidate of the 2.5 series.
1628 *(Released Tue, 4 Oct 2011)*
1630 This was the first release candidate of the 2.5 series.
1636 *(Released Wed, 31 Aug 2011)*
1638 This was the third beta release of the 2.5 series.
1644 *(Released Mon, 22 Aug 2011)*
1646 This was the second beta release of the 2.5 series.
1652 *(Released Fri, 12 Aug 2011)*
1654 This was the first beta release of the 2.5 series.
1660 *(Released Thu, 27 Oct 2011)*
1662 - Fixed bug when parsing command line parameter values ending in
1664 - Fixed assertion error after unclean master shutdown
1665 - Disable HTTP client pool for RPC, significantly reducing memory usage
1667 - Fixed queue archive creation with wrong permissions
1673 *(Released Tue, 23 Aug 2011)*
1677 - Fixed documentation for importing with ``--src-dir`` option
1678 - Fixed a bug in ``ensure-dirs`` with queue/archive permissions
1679 - Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
1685 *(Released Fri, 5 Aug 2011)*
1687 Many bug-fixes and a few small features:
1689 - Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
1690 issues when you tried to add an instance with two MAC addresses in one
1692 - KVM: fixed per-instance stored UID value
1693 - KVM: configure bridged NICs at migration start
1694 - KVM: Fix a bug where instance will not start with never KVM versions
1696 - Added OS search path to ``gnt-cluster info``
1697 - Fixed an issue with ``file_storage_dir`` where you were forced to
1698 provide an absolute path, but the documentation states it is a
1699 relative path, the documentation was right
1700 - Added a new parameter to instance stop/start called ``--no-remember``
1701 that will make the state change to not be remembered
1702 - Implemented ``no_remember`` at RAPI level
1703 - Improved the documentation
1704 - Node evacuation: don't call IAllocator if node is already empty
1705 - Fixed bug in DRBD8 replace disks on current nodes
1706 - Fixed bug in recreate-disks for DRBD instances
1707 - Moved assertion checking locks in ``gnt-instance replace-disks``
1708 causing it to abort with not owning the right locks for some situation
1709 - Job queue: Fixed potential race condition when cancelling queued jobs
1710 - Fixed off-by-one bug in job serial generation
1711 - ``gnt-node volumes``: Fix instance names
1712 - Fixed aliases in bash completion
1713 - Fixed a bug in reopening log files after being sent a SIGHUP
1714 - Added a flag to burnin to allow specifying VCPU count
1715 - Bugfixes to non-root Ganeti configuration
1721 *(Released Thu, 12 May 2011)*
1723 Many bug-fixes and a few new small features:
1725 - Fixed a bug related to log opening failures
1726 - Fixed a bug in instance listing with orphan instances
1727 - Fixed a bug which prevented resetting the cluster-level node parameter
1728 ``oob_program`` to the default
1729 - Many fixes related to the ``cluster-merge`` tool
1730 - Fixed a race condition in the lock monitor, which caused failures
1731 during (at least) creation of many instances in parallel
1732 - Improved output for gnt-job info
1733 - Removed the quiet flag on some ssh calls which prevented debugging
1735 - Improved the N+1 failure messages in cluster verify by actually
1736 showing the memory values (needed and available)
1737 - Increased lock attempt timeouts so that when executing long operations
1738 (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
1739 too early and thus prevent the use of the 'fair' mechanism
1740 - Changed instance query data (``gnt-instance info``) to not acquire
1741 locks unless needed, thus allowing its use on locked instance if only
1742 static information is asked for
1743 - Improved behaviour with filesystems that do not support rename on an
1745 - Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
1746 kept locks on all nodes during the wipe, which is unneeded
1747 - Fixed ``gnt-watcher`` handling of errors during hooks execution
1748 - Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
1749 10GiB) which caused the wipe to fail right at the end in some cases
1750 - Fixed master IP activation when doing master failover with no-voting
1751 - Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
1752 the master node itself
1753 - Fixed potential data-loss in under disk full conditions, where Ganeti
1754 wouldn't check correctly the return code and would consider
1755 partially-written files 'correct'
1756 - Fixed bug related to multiple VGs and DRBD disk replacing
1757 - Added new disk parameter ``metavg`` that allows placement of the meta
1758 device for DRBD in a different volume group
1759 - Fixed error handling in the node daemon when the system libc doesn't
1760 have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
1761 - Fixed lock release during replace-disks, which kept cluster-wide locks
1762 when doing disk replaces with an iallocator script
1763 - Added check for missing bridges in cluster verify
1764 - Handle EPIPE errors while writing to the terminal better, so that
1765 piping the output to e.g. ``less`` doesn't cause a backtrace
1766 - Fixed rare case where a ^C during Luxi calls could have been
1767 interpreted as server errors, instead of simply terminating
1768 - Fixed a race condition in LUGroupAssignNodes (``gnt-group
1770 - Added a few more parameters to the KVM hypervisor, allowing a second
1771 CDROM, custom disk type for CDROMs and a floppy image
1772 - Removed redundant message in instance rename when the name is given
1774 - Added option to ``gnt-instance recreate-disks`` to allow creating the
1775 disks on new nodes, allowing recreation when the original instance
1776 nodes are completely gone
1777 - Added option when converting disk templates to DRBD to skip waiting
1778 for the resync, in order to make the instance available sooner
1779 - Added two new variables to the OS scripts environment (containing the
1781 - Made the root_path and optional parameter for the xen-pvm hypervisor,
1782 to allow use of ``pvgrub`` as bootloader
1783 - Changed the instance memory modifications to only check out-of-memory
1784 conditions on memory increases, and turned the secondary node warnings
1785 into errors (they can still be overridden via ``--force``)
1786 - Fixed the handling of a corner case when the Python installation gets
1787 corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1788 to execute a command that doesn't exist
1789 - Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1790 node of the instance returned failures during instance shutdown; this
1791 adds the option ``--ignore-consistency`` to gnt-instance move
1793 And as usual, various improvements to the error messages, documentation
1800 *(Released Wed, 09 Mar 2011)*
1802 Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1803 the RAPI users file if run twice (even with ``--dry-run``).
1805 The release fixes that bug (nothing else changed).
1811 *(Released Mon, 07 Mar 2011)*
1813 Final 2.4.0 release. Just a few small fixes:
1815 - Fixed RAPI node evacuate
1816 - Fixed the kvm-ifup script
1817 - Fixed internal error handling for special job cases
1818 - Updated man page to specify the escaping feature for options
1824 *(Released Mon, 28 Feb 2011)*
1826 A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1827 that this feature wiped the disks of the wrong instance, leading to loss
1832 - Fixed title of query field containing instance name
1833 - Expanded the glossary in the documentation
1834 - Fixed one unittest (internal issue)
1840 *(Released Mon, 21 Feb 2011)*
1842 A number of bug fixes plus just a couple functionality changes.
1844 On the user-visible side, the ``gnt-* list`` command output has changed
1845 with respect to "special" field states. The current rc1 style of display
1846 can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1847 the default output mode special fields states are displayed as follows:
1849 - Offline resource: ``*``
1850 - Unavailable/not applicable: ``-``
1851 - Data missing (RPC failure): ``?``
1852 - Unknown field: ``??``
1854 Another user-visible change is the addition of ``--force-join`` to
1859 - ``tools/cluster-merge`` has seen many fixes and is now enabled again
1860 - Fixed regression in RAPI/instance reinstall where all parameters were
1861 required (instead of optional)
1862 - Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1863 - Fixed iallocator usage (offline nodes were not considered offline)
1864 - Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1865 - Fixed hypervisor and OS parameter validation with respect to
1866 non-vm_capable nodes
1867 - Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1869 - Fixed ``tools/listrunner`` with respect to agent-based usage
1875 *(Released Fri, 4 Feb 2011)*
1877 Many changes and fixes since the beta1 release. While there were some
1878 internal changes, the code has been mostly stabilised for the RC
1881 Note: the dumb allocator was removed in this release, as it was not kept
1882 up-to-date with the IAllocator protocol changes. It is recommended to
1883 use the ``hail`` command from the ganeti-htools package.
1885 Note: the 2.4 and up versions of Ganeti are not compatible with the
1886 0.2.x branch of ganeti-htools. You need to upgrade to
1887 ganeti-htools-0.3.0 (or later).
1889 Regressions fixed from 2.3
1890 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1892 - Fixed the ``gnt-cluster verify-disks`` command
1893 - Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1895 - Fixed disk adoption breakage
1896 - Fixed wrong headers in instance listing for field aliases
1901 - Fixed corner case in KVM handling of NICs
1902 - Fixed many cases of wrong handling of non-vm_capable nodes
1903 - Fixed a bug where a missing instance symlink was not possible to
1904 recreate with any ``gnt-*`` command (now ``gnt-instance
1905 activate-disks`` does it)
1906 - Fixed the volume group name as reported by ``gnt-cluster
1908 - Increased timeouts for the import-export code, hopefully leading to
1909 fewer aborts due network or instance timeouts
1910 - Fixed bug in ``gnt-node list-storage``
1911 - Fixed bug where not all daemons were started on cluster
1912 initialisation, but only at the first watcher run
1913 - Fixed many bugs in the OOB implementation
1914 - Fixed watcher behaviour in presence of instances with offline
1916 - Fixed instance list output for instances running on the wrong node
1917 - a few fixes to the cluster-merge tool, but it still cannot merge
1918 multi-node groups (currently it is not recommended to use this tool)
1924 - Improved network configuration for the KVM hypervisor
1925 - Added e1000 as a supported NIC for Xen-HVM
1926 - Improved the lvmstrap tool to also be able to use partitions, as
1927 opposed to full disks
1928 - Improved speed of disk wiping (the cluster parameter
1929 ``prealloc_wipe_disks``, so that it has a low impact on the total time
1930 of instance creations
1931 - Added documentation for the OS parameters
1932 - Changed ``gnt-instance deactivate-disks`` so that it can work if the
1933 hypervisor is not responding
1934 - Added display of blacklisted and hidden OS information in
1935 ``gnt-cluster info``
1936 - Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1937 NIC and node parameters, which might create problems with currently
1938 invalid (but undetected) configuration files, but prevents validation
1939 failures when unrelated parameters are modified
1940 - Changed cluster initialisation to wait for the master daemon to become
1942 - Expanded the RAPI interface:
1944 - Added config redistribution resource
1945 - Added activation/deactivation of instance disks
1946 - Added export of console information
1948 - Implemented log file reopening on SIGHUP, which allows using
1949 logrotate(8) for the Ganeti log files
1950 - Added a basic OOB helper script as an example
1956 *(Released Fri, 14 Jan 2011)*
1961 - Fixed timezone issues when formatting timestamps
1962 - Added support for node groups, available via ``gnt-group`` and other
1964 - Added out-of-band framework and management, see :doc:`design
1965 document <design-oob>`
1966 - Removed support for roman numbers from ``gnt-node list`` and
1967 ``gnt-instance list``.
1968 - Allowed modification of master network interface via ``gnt-cluster
1969 modify --master-netdev``
1970 - Accept offline secondaries while shutting down instance disks
1971 - Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1972 - Added support for multiple LVM volume groups
1973 - Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1975 - Added commands to list available fields:
1977 - ``gnt-node list-fields``
1978 - ``gnt-group list-fields``
1979 - ``gnt-instance list-fields``
1981 - Updated documentation and man pages
1986 - Moved ``rapi_users`` file into separate directory, now named
1987 ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1989 - Added new tool for running commands on many machines,
1990 ``tools/ganeti-listrunner``
1991 - Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1992 improving the ``gnt-instance console`` output
1993 - Allowed customisation of disk index separator at ``configure`` time
1994 - Export node group allocation policy to :doc:`iallocator <iallocator>`
1995 - Added support for non-partitioned md disks in ``lvmstrap``
1996 - Added script to gracefully power off KVM instances
1997 - Split ``utils`` module into smaller parts
1998 - Changed query operations to return more detailed information, e.g.
1999 whether an information is unavailable due to an offline node. To use
2000 this new functionality, the LUXI call ``Query`` must be used. Field
2001 information is now stored by the master daemon and can be retrieved
2002 using ``QueryFields``. Instances, nodes and groups can also be queried
2003 using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
2004 exposed via RAPI). The following commands make use of this
2005 infrastructure change:
2007 - ``gnt-group list``
2008 - ``gnt-group list-fields``
2010 - ``gnt-node list-fields``
2011 - ``gnt-instance list``
2012 - ``gnt-instance list-fields``
2013 - ``gnt-debug locks``
2018 - New RAPI resources (see :doc:`rapi`):
2022 - ``/2/groups/[group_name]``
2023 - ``/2/groups/[group_name]/assign-nodes``
2024 - ``/2/groups/[group_name]/modify``
2025 - ``/2/groups/[group_name]/rename``
2026 - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
2030 - Implemented ``no_install`` for instance creation
2031 - Implemented OS parameters for instance reinstallation, allowing
2032 use of special settings on reinstallation (e.g. for preserving data)
2037 - Added IPv6 support in import/export
2038 - Pause DRBD synchronization while wiping disks on instance creation
2039 - Updated unittests and QA scripts
2040 - Improved network parameters passed to KVM
2041 - Converted man pages from docbook to reStructuredText
2047 *(Released Mon, 20 Dec 2010)*
2049 Released version 2.3.1~rc1 without any changes.
2055 *(Released Wed, 1 Dec 2010)*
2057 - impexpd: Disable OpenSSL compression in socat if possible (backport
2058 from master, commit e90739d625b, see :doc:`installation guide
2059 <install-quick>` for details)
2060 - Changed unittest coverage report to exclude test scripts
2061 - Added script to check version format
2067 *(Released Wed, 1 Dec 2010)*
2069 Released version 2.3.0~rc1 without any changes.
2075 *(Released Fri, 19 Nov 2010)*
2077 A number of bugfixes and documentation updates:
2079 - Update ganeti-os-interface documentation
2080 - Fixed a bug related to duplicate MACs or similar items which should be
2082 - Fix breakage in OS state modify
2083 - Reinstall instance: disallow offline secondaries (fixes bug related to
2084 OS changing but reinstall failing)
2085 - plus all the other fixes between 2.2.1 and 2.2.2
2091 *(Released Tue, 2 Nov 2010)*
2093 - Fixed clearing of the default iallocator using ``gnt-cluster modify``
2094 - Fixed master failover race with watcher
2095 - Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
2097 - Accept previously stopped instance for export with instance removal
2098 - Simplify and extend the environment variables for instance OS scripts
2099 - Added new node flags, ``master_capable`` and ``vm_capable``
2100 - Added optional instance disk wiping prior during allocation. This is a
2101 cluster-wide option and can be set/modified using
2102 ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
2103 - Added IPv6 support, see :doc:`design document <design-2.3>` and
2104 :doc:`install-quick`
2105 - Added a new watcher option (``--ignore-pause``)
2106 - Added option to ignore offline node on instance start/stop
2107 (``--ignore-offline``)
2108 - Allow overriding OS parameters with ``gnt-instance reinstall``
2109 - Added ability to change node's secondary IP address using ``gnt-node
2111 - Implemented privilege separation for all daemons except
2112 ``ganeti-noded``, see ``configure`` options
2113 - Complain if an instance's disk is marked faulty in ``gnt-cluster
2115 - Implemented job priorities (see ``ganeti(7)`` manpage)
2116 - Ignore failures while shutting down instances during failover from
2118 - Exit daemon's bootstrap process only once daemon is ready
2119 - Export more information via ``LUInstanceQuery``/remote API
2120 - Improved documentation, QA and unittests
2121 - RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
2122 restart if the file was created or changed
2123 - Added LUXI protocol version sent with each request and response,
2124 allowing detection of server/client mismatches
2125 - Moved the Python scripts among gnt-* and ganeti-* into modules
2126 - Moved all code related to setting up SSH to an external script,
2128 - Infrastructure changes for node group support in future versions
2134 *(Released Fri, 19 Nov 2010)*
2136 A few small bugs fixed, and some improvements to the build system:
2138 - Fix documentation regarding conversion to drbd
2139 - Fix validation of parameters in cluster modify (``gnt-cluster modify
2141 - Fix error handling in node modify with multiple changes
2142 - Allow remote imports without checked names
2148 *(Released Tue, 19 Oct 2010)*
2150 - Disable SSL session ID cache in RPC client
2156 *(Released Thu, 14 Oct 2010)*
2158 - Fix interaction between Curl/GnuTLS and the Python's HTTP server
2159 (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
2161 - Fix problems with interaction between Curl and Python's HTTP server,
2162 resulting in increased speed in many RPC calls
2163 - Improve our release script to prevent breakage with older aclocal and
2170 *(Released Thu, 7 Oct 2010)*
2172 - Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
2173 value retrieved from master
2174 - Added support for blacklisted or hidden OS definitions
2175 - Added simple lock monitor (accessible via (``gnt-debug locks``)
2176 - Added support for -mem-path in KVM hypervisor abstraction layer
2177 - Allow overriding instance parameters in tool for inter-cluster
2178 instance moves (``tools/move-instance``)
2179 - Improved opcode summaries (e.g. in ``gnt-job list``)
2180 - Improve consistency of OS listing by sorting it
2181 - Documentation updates
2187 *(Released Fri, 8 Oct 2010)*
2189 - Rebuild with a newer autotools version, to fix python 2.6 compatibility
2195 *(Released Mon, 4 Oct 2010)*
2197 - Fixed regression in ``gnt-instance rename``
2203 *(Released Wed, 22 Sep 2010)*
2205 - Fixed OS_VARIANT variable for OS scripts
2206 - Fixed cluster tag operations via RAPI
2207 - Made ``setup-ssh`` exit with non-zero code if an error occurred
2208 - Disabled RAPI CA checks in watcher
2214 *(Released Mon, 23 Aug 2010)*
2216 - Support DRBD versions of the format "a.b.c.d"
2218 - Re-introduce support for usage from multiple threads in RAPI client
2219 - Instance renames and modify via RAPI
2220 - Work around race condition between processing and archival in job
2222 - Mark opcodes following failed one as failed, too
2223 - Job field ``lock_status`` was removed due to difficulties making it
2224 work with the changed job queue in Ganeti 2.2; a better way to monitor
2225 locks is expected for a later 2.2.x release
2226 - Fixed dry-run behaviour with many commands
2227 - Support ``ssh-agent`` again when adding nodes
2228 - Many additional bugfixes
2234 *(Released Fri, 30 Jul 2010)*
2236 Important change: the internal RPC mechanism between Ganeti nodes has
2237 changed from using a home-grown http library (based on the Python base
2238 libraries) to use the PycURL library. This requires that PycURL is
2239 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
2240 against GnuTLS by default. cURL's support for GnuTLS had known issues
2241 before cURL 7.21.0 and we recommend using the latest cURL release or
2242 linking against OpenSSL. Most other distributions already link PycURL
2243 and cURL against OpenSSL. The command::
2245 python -c 'import pycurl; print pycurl.version'
2247 can be used to determine the libraries PycURL and cURL are linked
2250 Other significant changes:
2252 - Rewrote much of the internals of the job queue, in order to achieve
2253 better parallelism; this decouples job query operations from the job
2254 processing, and it should allow much nicer behaviour of the master
2255 daemon under load, and it also has uncovered some long-standing bugs
2256 related to the job serialisation (now fixed)
2257 - Added a default iallocator setting to the cluster parameters,
2258 eliminating the need to always pass nodes or an iallocator for
2259 operations that require selection of new node(s)
2260 - Added experimental support for the LXC virtualization method
2261 - Added support for OS parameters, which allows the installation of
2262 instances to pass parameter to OS scripts in order to customise the
2264 - Added a hypervisor parameter controlling the migration type (live or
2265 non-live), since hypervisors have various levels of reliability; this
2266 has renamed the 'live' parameter to 'mode'
2267 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
2268 logical volumes, meaning that cluster verify will ignore them and not
2269 flag their presence as errors
2270 - The watcher will now reset the error count for failed instances after
2271 8 hours, thus allowing self-healing if the problem that caused the
2272 instances to be down/fail to start has cleared in the meantime
2273 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
2274 check for, and warn, if the drbd module parameter ``usermode_helper``
2275 is not consistent with the cluster-wide setting; this is needed to
2276 make diagnose easier of failed drbd creations
2277 - Started adding base IPv6 support, but this is not yet
2278 enabled/available for use
2279 - Rename operations (cluster, instance) will now return the new name,
2280 which is especially useful if a short name was passed in
2281 - Added support for instance migration in RAPI
2282 - Added a tool to pre-configure nodes for the SSH setup, before joining
2283 them to the cluster; this will allow in the future a simplified model
2284 for node joining (but not yet fully enabled in 2.2); this needs the
2285 paramiko python library
2286 - Fixed handling of name-resolving errors
2287 - Fixed consistency of job results on the error path
2288 - Fixed master-failover race condition when executed multiple times in
2290 - Fixed many bugs related to the job queue (mostly introduced during the
2291 2.2 development cycle, so not all are impacting 2.1)
2292 - Fixed instance migration with missing disk symlinks
2293 - Fixed handling of unknown jobs in ``gnt-job archive``
2294 - And many other small fixes/improvements
2298 - Enhanced both the unittest and the QA coverage
2299 - Switched the opcode validation to a generic model, and extended the
2300 validation to all opcode parameters
2301 - Changed more parts of the code that write shell scripts to use the
2303 - Switched the master daemon to use the asyncore library for the Luxi
2310 *(Released Thu, 17 Jun 2010)*
2312 - Added tool (``move-instance``) and infrastructure to move instances
2313 between separate clusters (see :doc:`separate documentation
2314 <move-instance>` and :doc:`design document <design-2.2>`)
2315 - Added per-request RPC timeout
2316 - RAPI now requires a Content-Type header for requests with a body (e.g.
2317 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
2318 :rfc:`2616` (HTTP/1.1), section 7.2.1)
2319 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
2321 - Implemented initial support for running Ganeti daemons as separate
2322 users, see configure-time flags ``--with-user-prefix`` and
2323 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
2325 - Instances can be removed after export (``gnt-backup export
2326 --remove-instance``)
2327 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
2328 key (instead of 1024 bit)
2329 - Added new cluster configuration file for cluster domain secret
2330 - Import/export now use SSL instead of SSH
2331 - Added support for showing estimated time when exporting an instance,
2332 see the ``ganeti-os-interface(7)`` manpage and look for
2339 *(Released Tue, 16 Nov 2010)*
2341 Some more bugfixes. Unless critical bugs occur, this will be the last
2344 - Fix case of MAC special-values
2345 - Fix mac checker regex
2346 - backend: Fix typo causing "out of range" error
2347 - Add missing --units in gnt-instance list man page
2353 *(Released Tue, 24 Aug 2010)*
2356 - Don't ignore secondary node silently on non-mirrored disk templates
2358 - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
2359 - Fix usb_mouse parameter breaking with vnc_console (issue 109)
2360 - Properly document the usb_mouse parameter
2361 - Fix path in ganeti-rapi(8) (issue 116)
2362 - Adjust error message when the ganeti user's .ssh directory is
2364 - Add same-node-check when changing the disk template to drbd
2370 *(Released Fri, 16 Jul 2010)*
2373 - Add an option to only select some reboot types during qa/burnin.
2374 (on some hypervisors consequent reboots are not supported)
2375 - Fix infrequent race condition in master failover. Sometimes the old
2376 master ip address would be still detected as up for a short time
2377 after it was removed, causing failover to fail.
2378 - Decrease mlockall warnings when the ctypes module is missing. On
2379 Python 2.4 we support running even if no ctypes module is installed,
2380 but we were too verbose about this issue.
2381 - Fix building on old distributions, on which man doesn't have a
2383 - Fix RAPI not to ignore the MAC address on instance creation
2384 - Implement the old instance creation format in the RAPI client.
2390 *(Released Thu, 01 Jul 2010)*
2392 A small bugfix release:
2393 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
2394 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
2395 a non-existing option
2396 - Fix instance create: the --force-variant option was ignored
2397 - Improve pylint 0.21 compatibility and warnings with Python 2.6
2398 - Fix modify node storage with non-FQDN arguments
2399 - Fix RAPI client to authenticate under Python 2.6 when used
2400 for more than 5 requests needing authentication
2401 - Fix gnt-instance modify -t (storage) giving a wrong error message
2402 when converting a non-shutdown drbd instance to plain
2408 *(Released Fri, 18 Jun 2010)*
2410 A small bugfix release:
2412 - Fix live migration of KVM instances started with older Ganeti
2413 versions which had fewer hypervisor parameters
2414 - Fix gnt-instance grow-disk on down instances
2415 - Fix an error-reporting bug during instance migration
2416 - Better checking of the ``--net`` and ``--disk`` values, to avoid
2417 silently ignoring broken ones
2418 - Fix an RPC error reporting bug affecting, for example, RAPI client
2420 - Fix bug triggered by different API version os-es on different nodes
2421 - Fix a bug in instance startup with custom hvparams: OS level
2422 parameters would fail to be applied.
2423 - Fix the RAPI client under Python 2.6 (but more work is needed to
2424 make it work completely well with OpenSSL)
2425 - Fix handling of errors when resolving names from DNS
2431 *(Released Thu, 3 Jun 2010)*
2433 A medium sized development cycle. Some new features, and some
2434 fixes/small improvements/cleanups.
2436 Significant features
2437 ~~~~~~~~~~~~~~~~~~~~
2439 The node deamon now tries to mlock itself into memory, unless the
2440 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
2441 its logs, and falls back to console logging. This allows emergency
2442 features such as ``gnt-node powercycle`` to work even in the event of a
2443 broken node disk (tested offlining the disk hosting the node's
2444 filesystem and dropping its memory caches; don't try this at home)
2446 KVM: add vhost-net acceleration support. It can be tested with a new
2447 enough version of the kernel and of qemu-kvm.
2449 KVM: Add instance chrooting feature. If you use privilege dropping for
2450 your VMs you can also now force them to chroot to an empty directory,
2451 before starting the emulated guest.
2453 KVM: Add maximum migration bandwith and maximum downtime tweaking
2454 support (requires a new-enough version of qemu-kvm).
2456 Cluster verify will now warn if the master node doesn't have the master
2457 ip configured on it.
2459 Add a new (incompatible) instance creation request format to RAPI which
2460 supports all parameters (previously only a subset was supported, and it
2461 wasn't possible to extend the old format to accomodate all the new
2462 features. The old format is still supported, and a client can check for
2463 this feature, before using it, by checking for its presence in the
2464 ``features`` RAPI resource.
2466 Now with ancient latin support. Try it passing the ``--roman`` option to
2467 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
2468 (requires the python-roman module to be installed, in order to work).
2473 As usual many internal code refactorings, documentation updates, and
2476 - Lots of improvements and cleanups to the experimental Remote API
2477 (RAPI) client library.
2478 - A new unit test suite for the core daemon libraries.
2479 - A fix to creating missing directories makes sure the umask is not
2480 applied anymore. This enforces the same directory permissions
2482 - Better handling terminating daemons with ctrl+c (used when running
2483 them in debugging mode).
2484 - Fix a race condition in live migrating a KVM instance, when stat()
2485 on the old proc status file returned EINVAL, which is an unexpected
2487 - Fixed manpage checking with newer man and utf-8 charachters. But now
2488 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
2494 *(Released Fri, 7 May 2010)*
2496 Fix a bug which prevented untagged KVM instances from starting.
2502 *(Released Fri, 7 May 2010)*
2504 Another release with a long development cycle, during which many
2505 different features were added.
2507 Significant features
2508 ~~~~~~~~~~~~~~~~~~~~
2510 The KVM hypervisor now can run the individual instances as non-root, to
2511 reduce the impact of a VM being hijacked due to bugs in the
2512 hypervisor. It is possible to run all instances as a single (non-root)
2513 user, to manually specify a user for each instance, or to dynamically
2514 allocate a user out of a cluster-wide pool to each instance, with the
2515 guarantee that no two instances will run under the same user ID on any
2518 An experimental RAPI client library, that can be used standalone
2519 (without the other Ganeti libraries), is provided in the source tree as
2520 ``lib/rapi/client.py``. Note this client might change its interface in
2521 the future, as we iterate on its capabilities.
2523 A new command, ``gnt-cluster renew-crypto`` has been added to easily
2524 replace the cluster's certificates and crypto keys. This might help in
2525 case they have been compromised, or have simply expired.
2527 A new disk option for instance creation has been added that allows one
2528 to "adopt" currently existing logical volumes, with data
2529 preservation. This should allow easier migration to Ganeti from
2530 unmanaged (or managed via other software) instances.
2532 Another disk improvement is the possibility to convert between redundant
2533 (DRBD) and plain (LVM) disk configuration for an instance. This should
2534 allow better scalability (starting with one node and growing the
2535 cluster, or shrinking a two-node cluster to one node).
2537 A new feature that could help with automated node failovers has been
2538 implemented: if a node sees itself as offline (by querying the master
2539 candidates), it will try to shutdown (hard) all instances and any active
2540 DRBD devices. This reduces the risk of duplicate instances if an
2541 external script automatically failovers the instances on such nodes. To
2542 enable this, the cluster parameter ``maintain_node_health`` should be
2543 enabled; in the future this option (per the name) will enable other
2544 automatic maintenance features.
2546 Instance export/import now will reuse the original instance
2547 specifications for all parameters; that means exporting an instance,
2548 deleting it and the importing it back should give an almost identical
2549 instance. Note that the default import behaviour has changed from
2550 before, where it created only one NIC; now it recreates the original
2553 Cluster verify has added a few new checks: SSL certificates validity,
2554 /etc/hosts consistency across the cluster, etc.
2559 As usual, many internal changes were done, documentation fixes,
2562 - Fixed cluster initialization with disabled cluster storage (regression
2563 introduced in 2.1.1)
2564 - File-based storage supports growing the disks
2565 - Fixed behaviour of node role changes
2566 - Fixed cluster verify for some corner cases, plus a general rewrite of
2567 cluster verify to allow future extension with more checks
2568 - Fixed log spamming by watcher and node daemon (regression introduced
2570 - Fixed possible validation issues when changing the list of enabled
2572 - Fixed cleanup of /etc/hosts during node removal
2573 - Fixed RAPI response for invalid methods
2574 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
2575 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
2576 from ide disks, etc.)
2577 - Allow OS changes without re-installation (to record a changed OS
2578 outside of Ganeti, or to allow OS renames)
2579 - Allow instance creation without OS installation (useful for example if
2580 the OS will be installed manually, or restored from a backup not in
2582 - Implemented option to make cluster ``copyfile`` use the replication
2584 - Added list of enabled hypervisors to ssconf (possibly useful for
2586 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
2588 - A partial form of node re-IP is possible via node readd, which now
2589 allows changed node primary IP
2590 - Command line utilities now show an informational message if the job is
2592 - The logs of the master daemon now show the PID/UID/GID of the
2599 *(Released Fri, 12 Mar 2010)*
2601 During the 2.1.0 long release candidate cycle, a lot of improvements and
2602 changes have accumulated with were released later as 2.1.1.
2607 The node evacuate command (``gnt-node evacuate``) was significantly
2608 rewritten, and as such the IAllocator protocol was changed - a new
2609 request type has been added. This unfortunate change during a stable
2610 series is designed to improve performance of node evacuations; on
2611 clusters with more than about five nodes and which are well-balanced,
2612 evacuation should proceed in parallel for all instances of the node
2613 being evacuated. As such, any existing IAllocator scripts need to be
2614 updated, otherwise the above command will fail due to the unknown
2615 request. The provided "dumb" allocator has not been updated; but the
2616 ganeti-htools package supports the new protocol since version 0.2.4.
2618 Another important change is increased validation of node and instance
2619 names. This might create problems in special cases, if invalid host
2620 names are being used.
2622 Also, a new layer of hypervisor parameters has been added, that sits at
2623 OS level between the cluster defaults and the instance ones. This allows
2624 customisation of virtualization parameters depending on the installed
2625 OS. For example instances with OS 'X' may have a different KVM kernel
2626 (or any other parameter) than the cluster defaults. This is intended to
2627 help managing a multiple OSes on the same cluster, without manual
2628 modification of each instance's parameters.
2630 A tool for merging clusters, ``cluster-merge``, has been added in the
2631 tools sub-directory.
2636 - Improved the int/float conversions that should make the code more
2637 robust in face of errors from the node daemons
2638 - Fixed the remove node code in case of internal configuration errors
2639 - Fixed the node daemon behaviour in face of inconsistent queue
2640 directory (e.g. read-only file-system where we can't open the files
2642 - Fixed the behaviour of gnt-node modify for master candidate demotion;
2643 now it either aborts cleanly or, if given the new "auto_promote"
2644 parameter, will automatically promote other nodes as needed
2645 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
2646 completely prevent Ganeti from working
2647 - Fixed bug for instance export when not all disks were successfully
2649 - Fixed behaviour of node add when the new node is slow in starting up
2651 - Fixed handling of signals in the LUXI client, which should improve
2652 behaviour of command-line scripts
2653 - Added checks for invalid node/instance names in the configuration (now
2654 flagged during cluster verify)
2655 - Fixed watcher behaviour for disk activation errors
2656 - Fixed two potentially endless loops in http library, which led to the
2657 RAPI daemon hanging and consuming 100% CPU in some cases
2658 - Fixed bug in RAPI daemon related to hashed passwords
2659 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
2661 - Enhanced compatibility with non-Debian OSes, but not using absolute
2662 path in some commands and allowing customisation of the ssh
2663 configuration directory
2664 - Fixed possible future issue with new Python versions by abiding to the
2665 proper use of ``__slots__`` attribute on classes
2666 - Added checks that should prevent directory traversal attacks
2667 - Many documentation fixes based on feedback from users
2672 - Added an "early_release" more for instance replace disks and node
2673 evacuate, where we release locks earlier and thus allow higher
2674 parallelism within the cluster
2675 - Added watcher hooks, intended to allow the watcher to restart other
2676 daemons (e.g. from the ganeti-nbma project), but they can be used of
2677 course for any other purpose
2678 - Added a compile-time disable for DRBD barriers, to increase
2679 performance if the administrator trusts the power supply or the
2680 storage system to not lose writes
2681 - Added the option of using syslog for logging instead of, or in
2682 addition to, Ganeti's own log files
2683 - Removed boot restriction for paravirtual NICs for KVM, recent versions
2684 can indeed boot from a paravirtual NIC
2685 - Added a generic debug level for many operations; while this is not
2686 used widely yet, it allows one to pass the debug value all the way to
2688 - Enhanced the hooks environment for instance moves (failovers,
2689 migrations) where the primary/secondary nodes changed during the
2690 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
2691 - Enhanced data validations for many user-supplied values; one important
2692 item is the restrictions imposed on instance and node names, which
2693 might reject some (invalid) host names
2694 - Add a configure-time option to disable file-based storage, if it's not
2695 needed; this allows greater security separation between the master
2696 node and the other nodes from the point of view of the inter-node RPC
2698 - Added user notification in interactive tools if job is waiting in the
2699 job queue or trying to acquire locks
2700 - Added log messages when a job is waiting for locks
2701 - Added filtering by node tags in instance operations which admit
2702 multiple instances (start, stop, reboot, reinstall)
2703 - Added a new tool for cluster mergers, ``cluster-merge``
2704 - Parameters from command line which are of the form ``a=b,c=d`` can now
2705 use backslash escapes to pass in values which contain commas,
2706 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
2708 - For KVM, the instance name is the first parameter passed to KVM, so
2709 that it's more visible in the process list
2715 *(Released Tue, 2 Mar 2010)*
2717 Ganeti 2.1 brings many improvements with it. Major changes:
2719 - Added infrastructure to ease automated disk repairs
2720 - Added new daemon to export configuration data in a cheaper way than
2721 using the remote API
2722 - Instance NICs can now be routed instead of being associated with a
2724 - Improved job locking logic to reduce impact of jobs acquiring multiple
2725 locks waiting for other long-running jobs
2727 In-depth implementation details can be found in the Ganeti 2.1 design
2733 - Added chroot hypervisor
2734 - Added more options to xen-hvm hypervisor (``kernel_path`` and
2736 - Added more options to xen-pvm hypervisor (``use_bootloader``,
2737 ``bootloader_path`` and ``bootloader_args``)
2738 - Added the ``use_localtime`` option for the xen-hvm and kvm
2739 hypervisors, and the default value for this has changed to false (in
2740 2.0 xen-hvm always enabled it)
2741 - Added luxi call to submit multiple jobs in one go
2742 - Added cluster initialization option to not modify ``/etc/hosts``
2744 - Added network interface parameters
2745 - Added dry run mode to some LUs
2746 - Added RAPI resources:
2748 - ``/2/instances/[instance_name]/info``
2749 - ``/2/instances/[instance_name]/replace-disks``
2750 - ``/2/nodes/[node_name]/evacuate``
2751 - ``/2/nodes/[node_name]/migrate``
2752 - ``/2/nodes/[node_name]/role``
2753 - ``/2/nodes/[node_name]/storage``
2754 - ``/2/nodes/[node_name]/storage/modify``
2755 - ``/2/nodes/[node_name]/storage/repair``
2757 - Added OpCodes to evacuate or migrate all instances on a node
2758 - Added new command to list storage elements on nodes (``gnt-node
2759 list-storage``) and modify them (``gnt-node modify-storage``)
2760 - Added new ssconf files with master candidate IP address
2761 (``ssconf_master_candidates_ips``), node primary IP address
2762 (``ssconf_node_primary_ips``) and node secondary IP address
2763 (``ssconf_node_secondary_ips``)
2764 - Added ``ganeti-confd`` and a client library to query the Ganeti
2765 configuration via UDP
2766 - Added ability to run hooks after cluster initialization and before
2768 - Added automatic mode for disk replace (``gnt-instance replace-disks
2770 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
2771 after catastrophic data-loss
2772 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2774 - Added ``gnt-instance move`` command to move instances
2775 - Added ``gnt-cluster watcher`` command to control watcher
2776 - Added ``gnt-node powercycle`` command to powercycle nodes
2777 - Added new job status field ``lock_status``
2778 - Added parseable error codes to cluster verification (``gnt-cluster
2779 verify --error-codes``) and made output less verbose (use
2780 ``--verbose`` to restore previous behaviour)
2781 - Added UUIDs to the main config entities (cluster, nodes, instances)
2782 - Added support for OS variants
2783 - Added support for hashed passwords in the Ganeti remote API users file
2785 - Added option to specify maximum timeout on instance shutdown
2786 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
2787 - Added new helper script to start and stop Ganeti daemons
2788 (``daemon-util``), with the intent to reduce the work necessary to
2789 adjust Ganeti for non-Debian distributions and to start/stop daemons
2791 - Added more unittests
2792 - Fixed critical bug in ganeti-masterd startup
2793 - Removed the configure-time ``kvm-migration-port`` parameter, this is
2794 now customisable at the cluster level for both the KVM and Xen
2795 hypervisors using the new ``migration_port`` parameter
2796 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2797 reinstalling an instance
2798 - Allowed ``@`` in tag names
2799 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2800 - Many documentation updates
2801 - Distribute hypervisor files on ``gnt-cluster redist-conf``
2802 - ``gnt-instance reinstall`` can now reinstall multiple instances
2803 - Updated many command line parameters
2804 - Introduced new OS API version 15
2805 - No longer support a default hypervisor
2806 - Treat virtual LVs as inexistent
2807 - Improved job locking logic to reduce lock contention
2808 - Match instance and node names case insensitively
2809 - Reimplemented bash completion script to be more complete
2816 *(Released Thu, 4 Feb 2010)*
2818 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2819 - Fix a string formatting bug
2820 - Improve safety of the code in some error paths
2821 - Improve data validation in the master of values returned from nodes
2827 *(Released Thu, 17 Dec 2009)*
2829 - Fix security issue due to missing validation of iallocator names; this
2830 allows local and remote execution of arbitrary executables
2831 - Fix failure of gnt-node list during instance removal
2832 - Ship the RAPI documentation in the archive
2838 *(Released Wed, 30 Sep 2009)*
2840 - Fixed many wrong messages
2841 - Fixed a few bugs related to the locking library
2842 - Fixed MAC checking at instance creation time
2843 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
2844 - Fixed a few issues related to signal handling in both daemons and
2846 - Fixed the example startup script provided
2847 - Fixed insserv dependencies in the example startup script (patch from
2849 - Fixed handling of drained nodes in the iallocator framework
2850 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2852 - Fixed error related to invalid job IDs in job polling
2853 - Fixed job/opcode persistence on unclean master shutdown
2854 - Fixed handling of partial job processing after unclean master
2856 - Fixed error reporting from LUs, previously all errors were converted
2857 into execution errors
2858 - Fixed error reporting from burnin
2859 - Decreased significantly the memory usage of the job queue
2860 - Optimised slightly multi-job submission
2861 - Optimised slightly opcode loading
2862 - Backported the multi-job submit framework from the development
2863 branch; multi-instance start and stop should be faster
2864 - Added script to clean archived jobs after 21 days; this will reduce
2865 the size of the queue directory
2866 - Added some extra checks in disk size tracking
2867 - Added an example ethers hook script
2868 - Added a cluster parameter that prevents Ganeti from modifying of
2870 - Added more node information to RAPI responses
2871 - Added a ``gnt-job watch`` command that allows following the ouput of a
2873 - Added a bind-address option to ganeti-rapi
2874 - Added more checks to the configuration verify
2875 - Enhanced the burnin script such that some operations can be retried
2877 - Converted instance reinstall to multi-instance model
2883 *(Released Fri, 7 Aug 2009)*
2885 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2886 to allow using the pre-2.0.2 behaviour in activation, if any existing
2887 instances have mismatched disk sizes in the configuration
2888 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
2889 any configuration mismatches for disk sizes
2890 - Added ``gnt-master cluste-failover --no-voting`` to allow master
2891 failover to work on two-node clusters
2892 - Fixed the ``--net`` option of ``gnt-backup import``, which was
2894 - Fixed detection of OS script errors in ``gnt-backup export``
2895 - Fixed exit code of ``gnt-backup export``
2901 *(Released Fri, 17 Jul 2009)*
2903 - Added experimental support for stripped logical volumes; this should
2904 enhance performance but comes with a higher complexity in the block
2905 device handling; stripping is only enabled when passing
2906 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2907 affected even in the non-stripped mode
2908 - Improved resiliency against transient failures at the end of DRBD
2909 resyncs, and in general of DRBD resync checks
2910 - Fixed a couple of issues with exports and snapshot errors
2911 - Fixed a couple of issues in instance listing
2912 - Added display of the disk size in ``gnt-instance info``
2913 - Fixed checking for valid OSes in instance creation
2914 - Fixed handling of the "vcpus" parameter in instance listing and in
2915 general of invalid parameters
2916 - Fixed http server library, and thus RAPI, to handle invalid
2917 username/password combinations correctly; this means that now they
2918 report unauthorized for queries too, not only for modifications,
2919 allowing earlier detect of configuration problems
2920 - Added a new "role" node list field, equivalent to the master/master
2921 candidate/drained/offline flags combinations
2922 - Fixed cluster modify and changes of candidate pool size
2923 - Fixed cluster verify error messages for wrong files on regular nodes
2924 - Fixed a couple of issues with node demotion from master candidate role
2925 - Fixed node readd issues
2926 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2927 - Added a new ``--no-voting`` option for masterfailover to fix failover
2928 on two-nodes clusters when the former master node is unreachable
2929 - Added instance reinstall over RAPI
2935 *(Released Tue, 16 Jun 2009)*
2937 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2938 allow re-adding the start in single-user option (regression from 1.2)
2939 - the watcher writes the instance status to a file, to allow monitoring
2940 to report the instance status (from the master) based on cached
2941 results of the watcher's queries; while this can get stale if the
2942 watcher is being locked due to other work on the cluster, this is
2943 still an improvement
2944 - the watcher now also restarts the node daemon and the rapi daemon if
2946 - fixed the watcher to handle full and drained queue cases
2947 - hooks export more instance data in the environment, which helps if
2948 hook scripts need to take action based on the instance's properties
2949 (no longer need to query back into ganeti)
2950 - instance failovers when the instance is stopped do not check for free
2951 RAM, so that failing over a stopped instance is possible in low memory
2953 - rapi uses queries for tags instead of jobs (for less job traffic), and
2954 for cluster tags it won't talk to masterd at all but read them from
2956 - a couple of error handling fixes in RAPI
2957 - drbd handling: improved the error handling of inconsistent disks after
2958 resync to reduce the frequency of "there are some degraded disks for
2959 this instance" messages
2960 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
2961 error handling path called a wrong function name)
2967 *(Released Wed, 27 May 2009)*
2969 - no changes from rc5
2975 *(Released Wed, 20 May 2009)*
2977 - fix a couple of bugs (validation, argument checks)
2978 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2979 - some small improvements to RAPI and IAllocator
2980 - make watcher automatically start the master daemon if down
2986 *(Released Mon, 27 Apr 2009)*
2988 - change the OS list to not require locks; this helps with big clusters
2989 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2990 volume group is broken
2991 - ``gnt-instance info``, without any arguments, doesn't run for all
2992 instances anymore; either pass ``--all`` or pass the desired
2993 instances; this helps against mistakes on big clusters where listing
2994 the information for all instances takes a long time
2995 - miscellaneous doc and man pages fixes
3001 *(Released Wed, 8 Apr 2009)*
3003 - Change the internal locking model of some ``gnt-node`` commands, in
3004 order to reduce contention (and blocking of master daemon) when
3005 batching many creation/reinstall jobs
3006 - Fixes to Xen soft reboot
3007 - No longer build documentation at build time, instead distribute it in
3008 the archive, in order to reduce the need for the whole docbook/rst
3015 *(Released Fri, 27 Mar 2009)*
3017 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
3018 - Fix watcher startup sequence, improves the behaviour of busy clusters
3019 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
3020 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
3021 bind address and other places
3022 - Some documentation fixes and updates
3028 *(Released Mon, 2 Mar 2009)*
3030 - More documentation updates, now all docs should be more-or-less
3032 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
3034 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
3036 - Fix an issue related to $libdir/run/ganeti and cluster creation
3042 *(Released Thu, 19 Feb 2009)*
3044 - Xen PVM and KVM have switched the default value for the instance root
3045 disk to the first partition on the first drive, instead of the whole
3046 drive; this means that the OS installation scripts must be changed
3048 - Man pages have been updated
3049 - RAPI has been switched by default to HTTPS, and the exported functions
3050 should all work correctly
3051 - RAPI v1 has been removed
3052 - Many improvements to the KVM hypervisor
3053 - Block device errors are now better reported
3054 - Many other bugfixes and small improvements
3060 *(Released Mon, 26 Jan 2009)*
3062 - Version 2 is a general rewrite of the code and therefore the
3063 differences are too many to list, see the design document for 2.0 in
3064 the ``doc/`` subdirectory for more details
3065 - In this beta version there is not yet a migration path from 1.2 (there
3066 will be one in the final 2.0 release)
3067 - A few significant changes are:
3069 - all commands are executed by a daemon (``ganeti-masterd``) and the
3070 various ``gnt-*`` commands are just front-ends to it
3071 - all the commands are entered into, and executed from a job queue,
3072 see the ``gnt-job(8)`` manpage
3073 - the RAPI daemon supports read-write operations, secured by basic
3074 HTTP authentication on top of HTTPS
3075 - DRBD version 0.7 support has been removed, DRBD 8 is the only
3076 supported version (when migrating from Ganeti 1.2 to 2.0, you need
3077 to migrate to DRBD 8 first while still running Ganeti 1.2)
3078 - DRBD devices are using statically allocated minor numbers, which
3079 will be assigned to existing instances during the migration process
3080 - there is support for both Xen PVM and Xen HVM instances running on
3082 - KVM virtualization is supported too
3083 - file-based storage has been implemented, which means that it is
3084 possible to run the cluster without LVM and DRBD storage, for
3085 example using a shared filesystem exported from shared storage (and
3086 still have live migration)
3092 *(Released Tue, 13 Jan 2009)*
3094 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
3095 - Reuse the old instance mac address by default on instance import, if
3096 the instance name is the same.
3097 - Handle situations in which the node info rpc returns incomplete
3099 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
3100 - Improved version of batcher:
3102 - state file support
3103 - instance mac address support
3104 - support for HVM clusters/instances
3106 - Add an option to show the number of cpu sockets and nodes in
3108 - Support OSes that handle more than one version of the OS api (but do
3109 not change the current API in any other way)
3110 - Fix ``gnt-node migrate``
3111 - ``gnt-debug`` man page
3112 - Fixes various more typos and small issues
3113 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
3119 *(Released Wed, 24 Sep 2008)*
3121 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
3122 type of disk exported to fully virtualized instances.
3123 - provide access to the serial console of HVM instances
3124 - instance auto_balance flag, set by default. If turned off it will
3125 avoid warnings on cluster verify if there is not enough memory to fail
3126 over an instance. in the future it will prevent automatically failing
3127 it over when we will support that.
3128 - batcher tool for instance creation, see ``tools/README.batcher``
3129 - ``gnt-instance reinstall --select-os`` to interactively select a new
3130 operating system when reinstalling an instance.
3131 - when changing the memory amount on instance modify a check has been
3132 added that the instance will be able to start. also warnings are
3133 emitted if the instance will not be able to fail over, if auto_balance
3135 - documentation fixes
3136 - sync fields between ``gnt-instance list/modify/add/import``
3137 - fix a race condition in drbd when the sync speed was set after giving
3138 the device a remote peer.
3144 *(Released Tue, 22 Jul 2008)*
3146 - note: the allowed size and number of tags per object were reduced
3147 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
3148 - fixed twisted 8.x compatibility
3149 - fixed ``gnt-instance replace-disks`` with iallocator
3150 - add TCP keepalives on twisted connections to detect restarted nodes
3151 - disk increase support, see ``gnt-instance grow-disk``
3152 - implement bulk node/instance query for RAPI
3153 - add tags in node/instance listing (optional)
3154 - experimental migration (and live migration) support, read the man page
3155 for ``gnt-instance migrate``
3156 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
3157 has some small improvements in handling its state file
3163 *(Released Fri, 13 Jun 2008)*
3165 - Experimental readonly, REST-based remote API implementation;
3166 automatically started on master node, TCP port 5080, if enabled by
3167 ``--enable-rapi`` parameter to configure script.
3168 - Instance allocator support. Add and import instance accept a
3169 ``--iallocator`` parameter, and call that instance allocator to decide
3170 which node to use for the instance. The iallocator document describes
3171 what's expected from an allocator script.
3172 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
3173 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
3174 node is lost there is still enough memory to fail over the instances
3176 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
3177 ``gnt-cluster verify``, to check for site-specific compliance. All the
3178 hooks will run, and their output, if any, will be displayed. Any
3179 failing hook will make the verification return an error value.
3180 - ``gnt-cluster verify`` now checks that its peers are reachable on the
3181 primary and secondary interfaces
3182 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
3183 that is still declared as part of the cluster and has failed.
3184 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
3185 specifying output fields, that just adds the chosen fields to the
3187 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
3188 export from the filesystem.
3189 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
3190 have been added. Using them you can enable/disable acpi and pae
3191 support, and specify a path for a cd image to be exported to the
3192 instance. These parameters as the name suggest only work on HVM
3194 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
3195 PAE support will be set to the previously hardcoded values, but the
3196 (previously hardcoded) path to the CDROM ISO image will be unset and
3197 if required, needs to be set manually with ``gnt-instance modify``
3199 - The address to which an instance's VNC console is bound is now
3200 selectable per-instance, rather than being cluster wide. Of course
3201 this only applies to instances controlled via VNC, so currently just
3202 applies to HVM clusters.
3208 *(Released Mon, 18 Feb 2008)*
3210 - more tweaks to the disk activation code (especially helpful for DRBD)
3211 - change the default ``gnt-instance list`` output format, now there is
3212 one combined status field (see the manpage for the exact values this
3214 - some more fixes for the mac export to hooks change
3215 - make Ganeti not break with DRBD 8.2.x (which changed the version
3216 format in ``/proc/drbd``) (issue 24)
3217 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
3218 template, allowing migration from DRBD0.7+MD to DRBD8
3224 *(Released Wed, 30 Jan 2008)*
3226 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
3228 - add command aliases infrastructure and a few aliases
3229 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
3230 man pages and the ``--help`` option of ``gnt-node
3231 list``/``gnt-instance list``
3232 - fix ``gnt-backup list`` with down nodes (issue 21)
3233 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
3234 - fix the dist archive and add a check for including svn/git files in
3236 - some developer-related changes: improve the burnin and the QA suite,
3237 add an upload script for testing during development
3243 *(Released Wed, 16 Jan 2008)*
3245 - experimental HVM support, read the install document, section
3246 "Initializing the cluster"
3247 - allow for the PVM hypervisor per-instance kernel and initrd paths
3248 - add a new command ``gnt-cluster verify-disks`` which uses a new
3249 algorithm to improve the reconnection of the DRBD pairs if the device
3250 on the secondary node has gone away
3251 - make logical volume code auto-activate LVs at disk activation time
3252 - slightly improve the speed of activating disks
3253 - allow specification of the MAC address at instance creation time, and
3254 changing it later via ``gnt-instance modify``
3255 - fix handling of external commands that generate lots of output on
3257 - update documentation with regard to minimum version of DRBD8 supported
3263 *(Released Tue, 4 Dec 2007)*
3265 - Log the ``xm create`` output to the node daemon log on failure (to
3266 help diagnosing the error)
3267 - In debug mode, log all external commands output if failed to the logs
3268 - Change parsing of lvm commands to ignore stderr
3274 *(Released Wed, 28 Nov 2007)*
3276 - Another round of updates to the DRBD 8 code to deal with more failures
3277 in the replace secondary node operation
3278 - Some more logging of failures in disk operations (lvm, drbd)
3279 - A few documentation updates
3286 *(Released Tue, 13 Nov 2007)*
3288 - Change configuration file format from Python's Pickle to JSON.
3289 Upgrading is possible using the cfgupgrade utility.
3290 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
3291 faster replace disks and is more stable (DRBD 8 has many improvements
3292 compared to DRBD 0.7)
3293 - Added command line tags support (see man pages for ``gnt-instance``,
3294 ``gnt-node``, ``gnt-cluster``)
3295 - Added instance rename support
3296 - Added multi-instance startup/shutdown
3297 - Added cluster rename support
3298 - Added ``gnt-node evacuate`` to simplify some node operations
3299 - Added instance reboot operation that can speedup reboot as compared to
3301 - Soften the requirement that hostnames are in FQDN format
3302 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
3304 - Removed dependency on debian's patched fping that uses the
3305 non-standard ``-S`` option
3306 - Now the OS definitions are searched for in multiple, configurable
3307 paths (easier for distros to package)
3308 - Some changes to the hooks infrastructure (especially the new
3309 post-configuration update hook)
3310 - Other small bugfixes
3312 .. vim: set textwidth=72 syntax=rst :