7 *(Released Wed, 22 Sep 2010)*
9 - Fixed OS_VARIANT variable for OS scripts
10 - Fixed cluster tag operations via RAPI
11 - Made ``setup-ssh`` exit with non-zero code if an error occurred
12 - Disabled RAPI CA checks in watcher
18 *(Released Mon, 23 Aug 2010)*
20 - Support DRBD versions of the format "a.b.c.d"
22 - Re-introduce support for usage from multiple threads in RAPI client
23 - Instance renames and modify via RAPI
24 - Work around race condition between processing and archival in job
26 - Mark opcodes following failed one as failed, too
27 - Job field ``lock_status`` was removed due to difficulties making it
28 work with the changed job queue in Ganeti 2.2; a better way to monitor
29 locks is expected for a later 2.2.x release
30 - Fixed dry-run behaviour with many commands
31 - Support ``ssh-agent`` again when adding nodes
32 - Many additional bugfixes
38 *(Released Fri, 30 Jul 2010)*
40 Important change: the internal RPC mechanism between Ganeti nodes has
41 changed from using a home-grown http library (based on the Python base
42 libraries) to use the PycURL library. This requires that PycURL is
43 installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
44 against GnuTLS by default. cURL's support for GnuTLS had known issues
45 before cURL 7.21.0 and we recommend using the latest cURL release or
46 linking against OpenSSL. Most other distributions already link PycURL
47 and cURL against OpenSSL. The command::
49 python -c 'import pycurl; print pycurl.version'
51 can be used to determine the libraries PycURL and cURL are linked
54 Other significant changes:
56 - Rewrote much of the internals of the job queue, in order to achieve
57 better parallelism; this decouples job query operations from the job
58 processing, and it should allow much nicer behaviour of the master
59 daemon under load, and it also has uncovered some long-standing bugs
60 related to the job serialisation (now fixed)
61 - Added a default iallocator setting to the cluster parameters,
62 eliminating the need to always pass nodes or an iallocator for
63 operations that require selection of new node(s)
64 - Added experimental support for the LXC virtualization method
65 - Added support for OS parameters, which allows the installation of
66 instances to pass parameter to OS scripts in order to customise the
68 - Added a hypervisor parameter controlling the migration type (live or
69 non-live), since hypervisors have various levels of reliability; this
70 has renamed the 'live' parameter to 'mode'
71 - Added a cluster parameter ``reserved_lvs`` that denotes reserved
72 logical volumes, meaning that cluster verify will ignore them and not
73 flag their presence as errors
74 - The watcher will now reset the error count for failed instances after
75 8 hours, thus allowing self-healing if the problem that caused the
76 instances to be down/fail to start has cleared in the meantime
77 - Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
78 check for, and warn, if the drbd module parameter ``usermode_helper``
79 is not consistent with the cluster-wide setting; this is needed to
80 make diagnose easier of failed drbd creations
81 - Started adding base IPv6 support, but this is not yet
82 enabled/available for use
83 - Rename operations (cluster, instance) will now return the new name,
84 which is especially useful if a short name was passed in
85 - Added support for instance migration in RAPI
86 - Added a tool to pre-configure nodes for the SSH setup, before joining
87 them to the cluster; this will allow in the future a simplified model
88 for node joining (but not yet fully enabled in 2.2); this needs the
89 paramiko python library
90 - Fixed handling of name-resolving errors
91 - Fixed consistency of job results on the error path
92 - Fixed master-failover race condition when executed multiple times in
94 - Fixed many bugs related to the job queue (mostly introduced during the
95 2.2 development cycle, so not all are impacting 2.1)
96 - Fixed instance migration with missing disk symlinks
97 - Fixed handling of unknown jobs in ``gnt-job archive``
98 - And many other small fixes/improvements
102 - Enhanced both the unittest and the QA coverage
103 - Switched the opcode validation to a generic model, and extended the
104 validation to all opcode parameters
105 - Changed more parts of the code that write shell scripts to use the
107 - Switched the master daemon to use the asyncore library for the Luxi
114 *(Released Thu, 17 Jun 2010)*
116 - Added tool (``move-instance``) and infrastructure to move instances
117 between separate clusters (see :doc:`separate documentation
118 <move-instance>` and :doc:`design document <design-2.2>`)
119 - Added per-request RPC timeout
120 - RAPI now requires a Content-Type header for requests with a body (e.g.
121 ``PUT`` or ``POST``) which must be set to ``application/json`` (see
122 :rfc:`2616` (HTTP/1.1), section 7.2.1)
123 - ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
125 - Implemented initial support for running Ganeti daemons as separate
126 users, see configure-time flags ``--with-user-prefix`` and
127 ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
129 - Instances can be removed after export (``gnt-backup export
131 - Self-signed certificates generated by Ganeti now use a 2048 bit RSA
132 key (instead of 1024 bit)
133 - Added new cluster configuration file for cluster domain secret
134 - Import/export now use SSL instead of SSH
135 - Added support for showing estimated time when exporting an instance,
136 see the ``ganeti-os-interface(7)`` manpage and look for
143 *(Released Fri, 16 Jul 2010)*
146 - Add an option to only select some reboot types during qa/burnin.
147 (on some hypervisors consequent reboots are not supported)
148 - Fix infrequent race condition in master failover. Sometimes the old
149 master ip address would be still detected as up for a short time
150 after it was removed, causing failover to fail.
151 - Decrease mlockall warnings when the ctypes module is missing. On
152 Python 2.4 we support running even if no ctypes module is installed,
153 but we were too verbose about this issue.
154 - Fix building on old distributions, on which man doesn't have a
156 - Fix RAPI not to ignore the MAC address on instance creation
157 - Implement the old instance creation format in the RAPI client.
163 *(Released Thu, 01 Jul 2010)*
165 A small bugfix release:
166 - Fix disk adoption: broken by strict --disk option checking in 2.1.4
167 - Fix batch-create: broken in the whole 2.1 series due to a lookup on
168 a non-existing option
169 - Fix instance create: the --force-variant option was ignored
170 - Improve pylint 0.21 compatibility and warnings with Python 2.6
171 - Fix modify node storage with non-FQDN arguments
172 - Fix RAPI client to authenticate under Python 2.6 when used
173 for more than 5 requests needing authentication
174 - Fix gnt-instance modify -t (storage) giving a wrong error message
175 when converting a non-shutdown drbd instance to plain
181 *(Released Fri, 18 Jun 2010)*
183 A small bugfix release:
185 - Fix live migration of KVM instances started with older Ganeti
186 versions which had fewer hypervisor parameters
187 - Fix gnt-instance grow-disk on down instances
188 - Fix an error-reporting bug during instance migration
189 - Better checking of the ``--net`` and ``--disk`` values, to avoid
190 silently ignoring broken ones
191 - Fix an RPC error reporting bug affecting, for example, RAPI client
193 - Fix bug triggered by different API version os-es on different nodes
194 - Fix a bug in instance startup with custom hvparams: OS level
195 parameters would fail to be applied.
196 - Fix the RAPI client under Python 2.6 (but more work is needed to
197 make it work completely well with OpenSSL)
198 - Fix handling of errors when resolving names from DNS
204 *(Released Thu, 3 Jun 2010)*
206 A medium sized development cycle. Some new features, and some
207 fixes/small improvements/cleanups.
212 The node deamon now tries to mlock itself into memory, unless the
213 ``--no-mlock`` flag is passed. It also doesn't fail if it can't write
214 its logs, and falls back to console logging. This allows emergency
215 features such as ``gnt-node powercycle`` to work even in the event of a
216 broken node disk (tested offlining the disk hosting the node's
217 filesystem and dropping its memory caches; don't try this at home)
219 KVM: add vhost-net acceleration support. It can be tested with a new
220 enough version of the kernel and of qemu-kvm.
222 KVM: Add instance chrooting feature. If you use privilege dropping for
223 your VMs you can also now force them to chroot to an empty directory,
224 before starting the emulated guest.
226 KVM: Add maximum migration bandwith and maximum downtime tweaking
227 support (requires a new-enough version of qemu-kvm).
229 Cluster verify will now warn if the master node doesn't have the master
232 Add a new (incompatible) instance creation request format to RAPI which
233 supports all parameters (previously only a subset was supported, and it
234 wasn't possible to extend the old format to accomodate all the new
235 features. The old format is still supported, and a client can check for
236 this feature, before using it, by checking for its presence in the
237 ``features`` RAPI resource.
239 Now with ancient latin support. Try it passing the ``--roman`` option to
240 ``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
241 (requires the python-roman module to be installed, in order to work).
246 As usual many internal code refactorings, documentation updates, and
249 - Lots of improvements and cleanups to the experimental Remote API
250 (RAPI) client library.
251 - A new unit test suite for the core daemon libraries.
252 - A fix to creating missing directories makes sure the umask is not
253 applied anymore. This enforces the same directory permissions
255 - Better handling terminating daemons with ctrl+c (used when running
256 them in debugging mode).
257 - Fix a race condition in live migrating a KVM instance, when stat()
258 on the old proc status file returned EINVAL, which is an unexpected
260 - Fixed manpage checking with newer man and utf-8 charachters. But now
261 you need the en_US.UTF-8 locale enabled to build Ganeti from git.
267 *(Released Fri, 7 May 2010)*
269 Fix a bug which prevented untagged KVM instances from starting.
275 *(Released Fri, 7 May 2010)*
277 Another release with a long development cycle, during which many
278 different features were added.
283 The KVM hypervisor now can run the individual instances as non-root, to
284 reduce the impact of a VM being hijacked due to bugs in the
285 hypervisor. It is possible to run all instances as a single (non-root)
286 user, to manually specify a user for each instance, or to dynamically
287 allocate a user out of a cluster-wide pool to each instance, with the
288 guarantee that no two instances will run under the same user ID on any
291 An experimental RAPI client library, that can be used standalone
292 (without the other Ganeti libraries), is provided in the source tree as
293 ``lib/rapi/client.py``. Note this client might change its interface in
294 the future, as we iterate on its capabilities.
296 A new command, ``gnt-cluster renew-crypto`` has been added to easily
297 replace the cluster's certificates and crypto keys. This might help in
298 case they have been compromised, or have simply expired.
300 A new disk option for instance creation has been added that allows one
301 to "adopt" currently existing logical volumes, with data
302 preservation. This should allow easier migration to Ganeti from
303 unmanaged (or managed via other software) instances.
305 Another disk improvement is the possibility to convert between redundant
306 (DRBD) and plain (LVM) disk configuration for an instance. This should
307 allow better scalability (starting with one node and growing the
308 cluster, or shrinking a two-node cluster to one node).
310 A new feature that could help with automated node failovers has been
311 implemented: if a node sees itself as offline (by querying the master
312 candidates), it will try to shutdown (hard) all instances and any active
313 DRBD devices. This reduces the risk of duplicate instances if an
314 external script automatically failovers the instances on such nodes. To
315 enable this, the cluster parameter ``maintain_node_health`` should be
316 enabled; in the future this option (per the name) will enable other
317 automatic maintenance features.
319 Instance export/import now will reuse the original instance
320 specifications for all parameters; that means exporting an instance,
321 deleting it and the importing it back should give an almost identical
322 instance. Note that the default import behaviour has changed from
323 before, where it created only one NIC; now it recreates the original
326 Cluster verify has added a few new checks: SSL certificates validity,
327 /etc/hosts consistency across the cluster, etc.
332 As usual, many internal changes were done, documentation fixes,
335 - Fixed cluster initialization with disabled cluster storage (regression
337 - File-based storage supports growing the disks
338 - Fixed behaviour of node role changes
339 - Fixed cluster verify for some corner cases, plus a general rewrite of
340 cluster verify to allow future extension with more checks
341 - Fixed log spamming by watcher and node daemon (regression introduced
343 - Fixed possible validation issues when changing the list of enabled
345 - Fixed cleanup of /etc/hosts during node removal
346 - Fixed RAPI response for invalid methods
347 - Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
348 - Multiple small improvements to the KVM hypervisor (VNC usage, booting
349 from ide disks, etc.)
350 - Allow OS changes without re-installation (to record a changed OS
351 outside of Ganeti, or to allow OS renames)
352 - Allow instance creation without OS installation (useful for example if
353 the OS will be installed manually, or restored from a backup not in
355 - Implemented option to make cluster ``copyfile`` use the replication
357 - Added list of enabled hypervisors to ssconf (possibly useful for
359 - Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
361 - A partial form of node re-IP is possible via node readd, which now
362 allows changed node primary IP
363 - Command line utilities now show an informational message if the job is
365 - The logs of the master daemon now show the PID/UID/GID of the
372 *(Released Fri, 12 Mar 2010)*
374 During the 2.1.0 long release candidate cycle, a lot of improvements and
375 changes have accumulated with were released later as 2.1.1.
380 The node evacuate command (``gnt-node evacuate``) was significantly
381 rewritten, and as such the IAllocator protocol was changed - a new
382 request type has been added. This unfortunate change during a stable
383 series is designed to improve performance of node evacuations; on
384 clusters with more than about five nodes and which are well-balanced,
385 evacuation should proceed in parallel for all instances of the node
386 being evacuated. As such, any existing IAllocator scripts need to be
387 updated, otherwise the above command will fail due to the unknown
388 request. The provided "dumb" allocator has not been updated; but the
389 ganeti-htools package supports the new protocol since version 0.2.4.
391 Another important change is increased validation of node and instance
392 names. This might create problems in special cases, if invalid host
393 names are being used.
395 Also, a new layer of hypervisor parameters has been added, that sits at
396 OS level between the cluster defaults and the instance ones. This allows
397 customisation of virtualization parameters depending on the installed
398 OS. For example instances with OS 'X' may have a different KVM kernel
399 (or any other parameter) than the cluster defaults. This is intended to
400 help managing a multiple OSes on the same cluster, without manual
401 modification of each instance's parameters.
403 A tool for merging clusters, ``cluster-merge``, has been added in the
409 - Improved the int/float conversions that should make the code more
410 robust in face of errors from the node daemons
411 - Fixed the remove node code in case of internal configuration errors
412 - Fixed the node daemon behaviour in face of inconsistent queue
413 directory (e.g. read-only file-system where we can't open the files
415 - Fixed the behaviour of gnt-node modify for master candidate demotion;
416 now it either aborts cleanly or, if given the new “auto_promote”
417 parameter, will automatically promote other nodes as needed
418 - Fixed compatibility with (unreleased yet) Python 2.6.5 that would
419 completely prevent Ganeti from working
420 - Fixed bug for instance export when not all disks were successfully
422 - Fixed behaviour of node add when the new node is slow in starting up
424 - Fixed handling of signals in the LUXI client, which should improve
425 behaviour of command-line scripts
426 - Added checks for invalid node/instance names in the configuration (now
427 flagged during cluster verify)
428 - Fixed watcher behaviour for disk activation errors
429 - Fixed two potentially endless loops in http library, which led to the
430 RAPI daemon hanging and consuming 100% CPU in some cases
431 - Fixed bug in RAPI daemon related to hashed passwords
432 - Fixed bug for unintended qemu-level bridging of multi-NIC KVM
434 - Enhanced compatibility with non-Debian OSes, but not using absolute
435 path in some commands and allowing customisation of the ssh
436 configuration directory
437 - Fixed possible future issue with new Python versions by abiding to the
438 proper use of ``__slots__`` attribute on classes
439 - Added checks that should prevent directory traversal attacks
440 - Many documentation fixes based on feedback from users
445 - Added an “early_release” more for instance replace disks and node
446 evacuate, where we release locks earlier and thus allow higher
447 parallelism within the cluster
448 - Added watcher hooks, intended to allow the watcher to restart other
449 daemons (e.g. from the ganeti-nbma project), but they can be used of
450 course for any other purpose
451 - Added a compile-time disable for DRBD barriers, to increase
452 performance if the administrator trusts the power supply or the
453 storage system to not lose writes
454 - Added the option of using syslog for logging instead of, or in
455 addition to, Ganeti's own log files
456 - Removed boot restriction for paravirtual NICs for KVM, recent versions
457 can indeed boot from a paravirtual NIC
458 - Added a generic debug level for many operations; while this is not
459 used widely yet, it allows one to pass the debug value all the way to
461 - Enhanced the hooks environment for instance moves (failovers,
462 migrations) where the primary/secondary nodes changed during the
463 operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
464 - Enhanced data validations for many user-supplied values; one important
465 item is the restrictions imposed on instance and node names, which
466 might reject some (invalid) host names
467 - Add a configure-time option to disable file-based storage, if it's not
468 needed; this allows greater security separation between the master
469 node and the other nodes from the point of view of the inter-node RPC
471 - Added user notification in interactive tools if job is waiting in the
472 job queue or trying to acquire locks
473 - Added log messages when a job is waiting for locks
474 - Added filtering by node tags in instance operations which admit
475 multiple instances (start, stop, reboot, reinstall)
476 - Added a new tool for cluster mergers, ``cluster-merge``
477 - Parameters from command line which are of the form ``a=b,c=d`` can now
478 use backslash escapes to pass in values which contain commas,
479 e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
481 - For KVM, the instance name is the first parameter passed to KVM, so
482 that it's more visible in the process list
488 *(Released Tue, 2 Mar 2010)*
490 Ganeti 2.1 brings many improvements with it. Major changes:
492 - Added infrastructure to ease automated disk repairs
493 - Added new daemon to export configuration data in a cheaper way than
495 - Instance NICs can now be routed instead of being associated with a
497 - Improved job locking logic to reduce impact of jobs acquiring multiple
498 locks waiting for other long-running jobs
500 In-depth implementation details can be found in the Ganeti 2.1 design
506 - Added chroot hypervisor
507 - Added more options to xen-hvm hypervisor (``kernel_path`` and
509 - Added more options to xen-pvm hypervisor (``use_bootloader``,
510 ``bootloader_path`` and ``bootloader_args``)
511 - Added the ``use_localtime`` option for the xen-hvm and kvm
512 hypervisors, and the default value for this has changed to false (in
513 2.0 xen-hvm always enabled it)
514 - Added luxi call to submit multiple jobs in one go
515 - Added cluster initialization option to not modify ``/etc/hosts``
517 - Added network interface parameters
518 - Added dry run mode to some LUs
519 - Added RAPI resources:
521 - ``/2/instances/[instance_name]/info``
522 - ``/2/instances/[instance_name]/replace-disks``
523 - ``/2/nodes/[node_name]/evacuate``
524 - ``/2/nodes/[node_name]/migrate``
525 - ``/2/nodes/[node_name]/role``
526 - ``/2/nodes/[node_name]/storage``
527 - ``/2/nodes/[node_name]/storage/modify``
528 - ``/2/nodes/[node_name]/storage/repair``
530 - Added OpCodes to evacuate or migrate all instances on a node
531 - Added new command to list storage elements on nodes (``gnt-node
532 list-storage``) and modify them (``gnt-node modify-storage``)
533 - Added new ssconf files with master candidate IP address
534 (``ssconf_master_candidates_ips``), node primary IP address
535 (``ssconf_node_primary_ips``) and node secondary IP address
536 (``ssconf_node_secondary_ips``)
537 - Added ``ganeti-confd`` and a client library to query the Ganeti
538 configuration via UDP
539 - Added ability to run hooks after cluster initialization and before
541 - Added automatic mode for disk replace (``gnt-instance replace-disks
543 - Added ``gnt-instance recreate-disks`` to re-create (empty) disks
544 after catastrophic data-loss
545 - Added ``gnt-node repair-storage`` command to repair damaged LVM volume
547 - Added ``gnt-instance move`` command to move instances
548 - Added ``gnt-cluster watcher`` command to control watcher
549 - Added ``gnt-node powercycle`` command to powercycle nodes
550 - Added new job status field ``lock_status``
551 - Added parseable error codes to cluster verification (``gnt-cluster
552 verify --error-codes``) and made output less verbose (use
553 ``--verbose`` to restore previous behaviour)
554 - Added UUIDs to the main config entities (cluster, nodes, instances)
555 - Added support for OS variants
556 - Added support for hashed passwords in the Ganeti remote API users file
558 - Added option to specify maximum timeout on instance shutdown
559 - Added ``--no-ssh-init`` option to ``gnt-cluster init``
560 - Added new helper script to start and stop Ganeti daemons
561 (``daemon-util``), with the intent to reduce the work necessary to
562 adjust Ganeti for non-Debian distributions and to start/stop daemons
564 - Added more unittests
565 - Fixed critical bug in ganeti-masterd startup
566 - Removed the configure-time ``kvm-migration-port`` parameter, this is
567 now customisable at the cluster level for both the KVM and Xen
568 hypervisors using the new ``migration_port`` parameter
569 - Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
570 reinstalling an instance
571 - Allowed ``@`` in tag names
572 - Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
573 - Many documentation updates
574 - Distribute hypervisor files on ``gnt-cluster redist-conf``
575 - ``gnt-instance reinstall`` can now reinstall multiple instances
576 - Updated many command line parameters
577 - Introduced new OS API version 15
578 - No longer support a default hypervisor
579 - Treat virtual LVs as inexistent
580 - Improved job locking logic to reduce lock contention
581 - Match instance and node names case insensitively
582 - Reimplemented bash completion script to be more complete
589 *(Released Thu, 4 Feb 2010)*
591 - Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
592 - Fix a string formatting bug
593 - Improve safety of the code in some error paths
594 - Improve data validation in the master of values returned from nodes
600 *(Released Thu, 17 Dec 2009)*
602 - Fix security issue due to missing validation of iallocator names; this
603 allows local and remote execution of arbitrary executables
604 - Fix failure of gnt-node list during instance removal
605 - Ship the RAPI documentation in the archive
611 *(Released Wed, 30 Sep 2009)*
613 - Fixed many wrong messages
614 - Fixed a few bugs related to the locking library
615 - Fixed MAC checking at instance creation time
616 - Fixed a DRBD parsing bug related to gaps in /proc/drbd
617 - Fixed a few issues related to signal handling in both daemons and
619 - Fixed the example startup script provided
620 - Fixed insserv dependencies in the example startup script (patch from
622 - Fixed handling of drained nodes in the iallocator framework
623 - Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
625 - Fixed error related to invalid job IDs in job polling
626 - Fixed job/opcode persistence on unclean master shutdown
627 - Fixed handling of partial job processing after unclean master
629 - Fixed error reporting from LUs, previously all errors were converted
630 into execution errors
631 - Fixed error reporting from burnin
632 - Decreased significantly the memory usage of the job queue
633 - Optimised slightly multi-job submission
634 - Optimised slightly opcode loading
635 - Backported the multi-job submit framework from the development
636 branch; multi-instance start and stop should be faster
637 - Added script to clean archived jobs after 21 days; this will reduce
638 the size of the queue directory
639 - Added some extra checks in disk size tracking
640 - Added an example ethers hook script
641 - Added a cluster parameter that prevents Ganeti from modifying of
643 - Added more node information to RAPI responses
644 - Added a ``gnt-job watch`` command that allows following the ouput of a
646 - Added a bind-address option to ganeti-rapi
647 - Added more checks to the configuration verify
648 - Enhanced the burnin script such that some operations can be retried
650 - Converted instance reinstall to multi-instance model
656 *(Released Fri, 7 Aug 2009)*
658 - Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
659 to allow using the pre-2.0.2 behaviour in activation, if any existing
660 instances have mismatched disk sizes in the configuration
661 - Added ``gnt-cluster repair-disk-sizes`` command to check and update
662 any configuration mismatches for disk sizes
663 - Added ``gnt-master cluste-failover --no-voting`` to allow master
664 failover to work on two-node clusters
665 - Fixed the ``--net`` option of ``gnt-backup import``, which was
667 - Fixed detection of OS script errors in ``gnt-backup export``
668 - Fixed exit code of ``gnt-backup export``
674 *(Released Fri, 17 Jul 2009)*
676 - Added experimental support for stripped logical volumes; this should
677 enhance performance but comes with a higher complexity in the block
678 device handling; stripping is only enabled when passing
679 ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
680 affected even in the non-stripped mode
681 - Improved resiliency against transient failures at the end of DRBD
682 resyncs, and in general of DRBD resync checks
683 - Fixed a couple of issues with exports and snapshot errors
684 - Fixed a couple of issues in instance listing
685 - Added display of the disk size in ``gnt-instance info``
686 - Fixed checking for valid OSes in instance creation
687 - Fixed handling of the "vcpus" parameter in instance listing and in
688 general of invalid parameters
689 - Fixed http server library, and thus RAPI, to handle invalid
690 username/password combinations correctly; this means that now they
691 report unauthorized for queries too, not only for modifications,
692 allowing earlier detect of configuration problems
693 - Added a new "role" node list field, equivalent to the master/master
694 candidate/drained/offline flags combinations
695 - Fixed cluster modify and changes of candidate pool size
696 - Fixed cluster verify error messages for wrong files on regular nodes
697 - Fixed a couple of issues with node demotion from master candidate role
698 - Fixed node readd issues
699 - Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
700 - Added a new ``--no-voting`` option for masterfailover to fix failover
701 on two-nodes clusters when the former master node is unreachable
702 - Added instance reinstall over RAPI
708 *(Released Tue, 16 Jun 2009)*
710 - added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
711 allow re-adding the start in single-user option (regression from 1.2)
712 - the watcher writes the instance status to a file, to allow monitoring
713 to report the instance status (from the master) based on cached
714 results of the watcher's queries; while this can get stale if the
715 watcher is being locked due to other work on the cluster, this is
717 - the watcher now also restarts the node daemon and the rapi daemon if
719 - fixed the watcher to handle full and drained queue cases
720 - hooks export more instance data in the environment, which helps if
721 hook scripts need to take action based on the instance's properties
722 (no longer need to query back into ganeti)
723 - instance failovers when the instance is stopped do not check for free
724 RAM, so that failing over a stopped instance is possible in low memory
726 - rapi uses queries for tags instead of jobs (for less job traffic), and
727 for cluster tags it won't talk to masterd at all but read them from
729 - a couple of error handling fixes in RAPI
730 - drbd handling: improved the error handling of inconsistent disks after
731 resync to reduce the frequency of "there are some degraded disks for
732 this instance" messages
733 - fixed a bug in live migration when DRBD doesn't want to reconnect (the
734 error handling path called a wrong function name)
740 *(Released Wed, 27 May 2009)*
742 - no changes from rc5
745 Version 2.0 release candidate 5
746 -------------------------------
748 *(Released Wed, 20 May 2009)*
750 - fix a couple of bugs (validation, argument checks)
751 - fix ``gnt-cluster getmaster`` on non-master nodes (regression)
752 - some small improvements to RAPI and IAllocator
753 - make watcher automatically start the master daemon if down
756 Version 2.0 release candidate 4
757 -------------------------------
759 *(Released Mon, 27 Apr 2009)*
761 - change the OS list to not require locks; this helps with big clusters
762 - fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
763 volume group is broken
764 - ``gnt-instance info``, without any arguments, doesn't run for all
765 instances anymore; either pass ``--all`` or pass the desired
766 instances; this helps against mistakes on big clusters where listing
767 the information for all instances takes a long time
768 - miscellaneous doc and man pages fixes
771 Version 2.0 release candidate 3
772 -------------------------------
774 *(Released Wed, 8 Apr 2009)*
776 - Change the internal locking model of some ``gnt-node`` commands, in
777 order to reduce contention (and blocking of master daemon) when
778 batching many creation/reinstall jobs
779 - Fixes to Xen soft reboot
780 - No longer build documentation at build time, instead distribute it in
781 the archive, in order to reduce the need for the whole docbook/rst
785 Version 2.0 release candidate 2
786 -------------------------------
788 *(Released Fri, 27 Mar 2009)*
790 - Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
791 - Fix watcher startup sequence, improves the behaviour of busy clusters
792 - Some other fixes in ``gnt-cluster verify``, ``gnt-instance
793 replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
794 bind address and other places
795 - Some documentation fixes and updates
798 Version 2.0 release candidate 1
799 -------------------------------
801 *(Released Mon, 2 Mar 2009)*
803 - More documentation updates, now all docs should be more-or-less
805 - A couple of small fixes (mixed hypervisor clusters, offline nodes,
807 - Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
809 - Fix an issue related to $libdir/run/ganeti and cluster creation
815 *(Released Thu, 19 Feb 2009)*
817 - Xen PVM and KVM have switched the default value for the instance root
818 disk to the first partition on the first drive, instead of the whole
819 drive; this means that the OS installation scripts must be changed
821 - Man pages have been updated
822 - RAPI has been switched by default to HTTPS, and the exported functions
823 should all work correctly
824 - RAPI v1 has been removed
825 - Many improvements to the KVM hypervisor
826 - Block device errors are now better reported
827 - Many other bugfixes and small improvements
833 *(Released Mon, 26 Jan 2009)*
835 - Version 2 is a general rewrite of the code and therefore the
836 differences are too many to list, see the design document for 2.0 in
837 the ``doc/`` subdirectory for more details
838 - In this beta version there is not yet a migration path from 1.2 (there
839 will be one in the final 2.0 release)
840 - A few significant changes are:
842 - all commands are executed by a daemon (``ganeti-masterd``) and the
843 various ``gnt-*`` commands are just front-ends to it
844 - all the commands are entered into, and executed from a job queue,
845 see the ``gnt-job(8)`` manpage
846 - the RAPI daemon supports read-write operations, secured by basic
847 HTTP authentication on top of HTTPS
848 - DRBD version 0.7 support has been removed, DRBD 8 is the only
849 supported version (when migrating from Ganeti 1.2 to 2.0, you need
850 to migrate to DRBD 8 first while still running Ganeti 1.2)
851 - DRBD devices are using statically allocated minor numbers, which
852 will be assigned to existing instances during the migration process
853 - there is support for both Xen PVM and Xen HVM instances running on
855 - KVM virtualization is supported too
856 - file-based storage has been implemented, which means that it is
857 possible to run the cluster without LVM and DRBD storage, for
858 example using a shared filesystem exported from shared storage (and
859 still have live migration)
865 *(Released Tue, 13 Jan 2009)*
867 - Change the default reboot type in ``gnt-instance reboot`` to "hard"
868 - Reuse the old instance mac address by default on instance import, if
869 the instance name is the same.
870 - Handle situations in which the node info rpc returns incomplete
872 - Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
873 - Improved version of batcher:
876 - instance mac address support
877 - support for HVM clusters/instances
879 - Add an option to show the number of cpu sockets and nodes in
881 - Support OSes that handle more than one version of the OS api (but do
882 not change the current API in any other way)
883 - Fix ``gnt-node migrate``
884 - ``gnt-debug`` man page
885 - Fixes various more typos and small issues
886 - Increase disk resync maximum speed to 60MB/s (from 30MB/s)
892 *(Released Wed, 24 Sep 2008)*
894 - new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
895 type of disk exported to fully virtualized instances.
896 - provide access to the serial console of HVM instances
897 - instance auto_balance flag, set by default. If turned off it will
898 avoid warnings on cluster verify if there is not enough memory to fail
899 over an instance. in the future it will prevent automatically failing
900 it over when we will support that.
901 - batcher tool for instance creation, see ``tools/README.batcher``
902 - ``gnt-instance reinstall --select-os`` to interactively select a new
903 operating system when reinstalling an instance.
904 - when changing the memory amount on instance modify a check has been
905 added that the instance will be able to start. also warnings are
906 emitted if the instance will not be able to fail over, if auto_balance
908 - documentation fixes
909 - sync fields between ``gnt-instance list/modify/add/import``
910 - fix a race condition in drbd when the sync speed was set after giving
911 the device a remote peer.
917 *(Released Tue, 22 Jul 2008)*
919 - note: the allowed size and number of tags per object were reduced
920 - fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
921 - fixed twisted 8.x compatibility
922 - fixed ``gnt-instance replace-disks`` with iallocator
923 - add TCP keepalives on twisted connections to detect restarted nodes
924 - disk increase support, see ``gnt-instance grow-disk``
925 - implement bulk node/instance query for RAPI
926 - add tags in node/instance listing (optional)
927 - experimental migration (and live migration) support, read the man page
928 for ``gnt-instance migrate``
929 - the ``ganeti-watcher`` logs are now timestamped, and the watcher also
930 has some small improvements in handling its state file
936 *(Released Fri, 13 Jun 2008)*
938 - Experimental readonly, REST-based remote API implementation;
939 automatically started on master node, TCP port 5080, if enabled by
940 ``--enable-rapi`` parameter to configure script.
941 - Instance allocator support. Add and import instance accept a
942 ``--iallocator`` parameter, and call that instance allocator to decide
943 which node to use for the instance. The iallocator document describes
944 what's expected from an allocator script.
945 - ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
946 ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
947 node is lost there is still enough memory to fail over the instances
949 - ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
950 ``gnt-cluster verify``, to check for site-specific compliance. All the
951 hooks will run, and their output, if any, will be displayed. Any
952 failing hook will make the verification return an error value.
953 - ``gnt-cluster verify`` now checks that its peers are reachable on the
954 primary and secondary interfaces
955 - ``gnt-node add`` now supports the ``--readd`` option, to readd a node
956 that is still declared as part of the cluster and has failed.
957 - ``gnt-* list`` commands now accept a new ``-o +field`` way of
958 specifying output fields, that just adds the chosen fields to the
960 - ``gnt-backup`` now has a new ``remove`` command to delete an existing
961 export from the filesystem.
962 - New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
963 have been added. Using them you can enable/disable acpi and pae
964 support, and specify a path for a cd image to be exported to the
965 instance. These parameters as the name suggest only work on HVM
967 - When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
968 PAE support will be set to the previously hardcoded values, but the
969 (previously hardcoded) path to the CDROM ISO image will be unset and
970 if required, needs to be set manually with ``gnt-instance modify``
972 - The address to which an instance's VNC console is bound is now
973 selectable per-instance, rather than being cluster wide. Of course
974 this only applies to instances controlled via VNC, so currently just
975 applies to HVM clusters.
981 *(Released Mon, 18 Feb 2008)*
983 - more tweaks to the disk activation code (especially helpful for DRBD)
984 - change the default ``gnt-instance list`` output format, now there is
985 one combined status field (see the manpage for the exact values this
987 - some more fixes for the mac export to hooks change
988 - make Ganeti not break with DRBD 8.2.x (which changed the version
989 format in ``/proc/drbd``) (issue 24)
990 - add an upgrade tool from "remote_raid1" disk template to "drbd" disk
991 template, allowing migration from DRBD0.7+MD to DRBD8
997 *(Released Wed, 30 Jan 2008)*
999 - fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
1001 - add command aliases infrastructure and a few aliases
1002 - allow listing of VCPUs in the ``gnt-instance list`` and improve the
1003 man pages and the ``--help`` option of ``gnt-node
1004 list``/``gnt-instance list``
1005 - fix ``gnt-backup list`` with down nodes (issue 21)
1006 - change the tools location (move from $pkgdatadir to $pkglibdir/tools)
1007 - fix the dist archive and add a check for including svn/git files in
1009 - some developer-related changes: improve the burnin and the QA suite,
1010 add an upload script for testing during development
1016 *(Released Wed, 16 Jan 2008)*
1018 - experimental HVM support, read the install document, section
1019 "Initializing the cluster"
1020 - allow for the PVM hypervisor per-instance kernel and initrd paths
1021 - add a new command ``gnt-cluster verify-disks`` which uses a new
1022 algorithm to improve the reconnection of the DRBD pairs if the device
1023 on the secondary node has gone away
1024 - make logical volume code auto-activate LVs at disk activation time
1025 - slightly improve the speed of activating disks
1026 - allow specification of the MAC address at instance creation time, and
1027 changing it later via ``gnt-instance modify``
1028 - fix handling of external commands that generate lots of output on
1030 - update documentation with regard to minimum version of DRBD8 supported
1036 *(Released Tue, 4 Dec 2007)*
1038 - Log the ``xm create`` output to the node daemon log on failure (to
1039 help diagnosing the error)
1040 - In debug mode, log all external commands output if failed to the logs
1041 - Change parsing of lvm commands to ignore stderr
1047 *(Released Wed, 28 Nov 2007)*
1049 - Another round of updates to the DRBD 8 code to deal with more failures
1050 in the replace secondary node operation
1051 - Some more logging of failures in disk operations (lvm, drbd)
1052 - A few documentation updates
1059 *(Released Tue, 13 Nov 2007)*
1061 - Change configuration file format from Python's Pickle to JSON.
1062 Upgrading is possible using the cfgupgrade utility.
1063 - Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
1064 faster replace disks and is more stable (DRBD 8 has many improvements
1065 compared to DRBD 0.7)
1066 - Added command line tags support (see man pages for ``gnt-instance``,
1067 ``gnt-node``, ``gnt-cluster``)
1068 - Added instance rename support
1069 - Added multi-instance startup/shutdown
1070 - Added cluster rename support
1071 - Added ``gnt-node evacuate`` to simplify some node operations
1072 - Added instance reboot operation that can speedup reboot as compared to
1074 - Soften the requirement that hostnames are in FQDN format
1075 - The ``ganeti-watcher`` now activates drbd pairs after secondary node
1077 - Removed dependency on debian's patched fping that uses the
1078 non-standard ``-S`` option
1079 - Now the OS definitions are searched for in multiple, configurable
1080 paths (easier for distros to package)
1081 - Some changes to the hooks infrastructure (especially the new
1082 post-configuration update hook)
1083 - Other small bugfixes
1085 .. vim: set textwidth=72 :