Statistics
| Branch: | Tag: | Revision:

root / NEWS @ 2d6bdcc5

History | View | Annotate | Download (96.2 kB)

1
News
2
====
3

    
4

    
5
Version 2.7.0 beta1
6
-------------------
7

    
8
*(Released Wed, 6 Feb 2013)*
9

    
10
- ``gnt-instance batch-create`` has been changed to use the bulk create
11
  opcode from Ganeti. This lead to incompatible changes in the format of
12
  the JSON file. It's now not a custom dict anymore but a dict
13
  compatible with the ``OpInstanceCreate`` opcode.
14
- Parent directories for file storage need to be listed in
15
  ``$sysconfdir/ganeti/file-storage-paths`` now. ``cfgupgrade`` will
16
  write the file automatically based on old configuration values, but it
17
  can not distribute it across all nodes and the file contents should be
18
  verified. Use ``gnt-cluster copyfile
19
  $sysconfdir/ganeti/file-storage-paths`` once the cluster has been
20
  upgraded. The reason for requiring this list of paths now is that
21
  before it would have been possible to inject new paths via RPC,
22
  allowing files to be created in arbitrary locations. The RPC protocol
23
  is protected using SSL/X.509 certificates, but as a design principle
24
  Ganeti does not permit arbitrary paths to be passed.
25
- The parsing of the variants file for OSes (see
26
  :manpage:`ganeti-os-interface(7)`) has been slightly changed: now empty
27
  lines and comment lines (starting with ``#``) are ignored for better
28
  readability.
29
- The ``setup-ssh`` tool added in Ganeti 2.2 has been replaced and is no
30
  longer available. ``gnt-node add`` now invokes a new tool on the
31
  destination node, named ``prepare-node-join``, to configure the SSH
32
  daemon. Paramiko is no longer necessary to configure nodes' SSH
33
  daemons via ``gnt-node add``.
34
- A new user option, :pyeval:`rapi.RAPI_ACCESS_READ`, has been added
35
  for RAPI users. It allows granting permissions to query for
36
  information to a specific user without giving
37
  :pyeval:`rapi.RAPI_ACCESS_WRITE` permissions.
38
- A new tool named ``node-cleanup`` has been added. It cleans remains of
39
  a cluster from a machine by stopping all daemons, removing
40
  certificates and ssconf files. Unless the ``--no-backup`` option is
41
  given, copies of the certificates are made.
42
- Draining (``gnt-cluster queue drain``) and un-draining the job queue
43
  (``gnt-cluster queue undrain``) now affects all nodes in a cluster and
44
  the flag is not reset after a master failover.
45
- Instance creations now support the use of opportunistic locking,
46
  potentially speeding up the (parallel) creation of multiple instances.
47
  This feature is currently only available via the :doc:`RAPI
48
  <rapi>` interface and when an instance allocator is used. If the
49
  ``opportunistic_locking`` parameter is set the opcode will try to
50
  acquire as many locks as possible, but will not wait for any locks
51
  held by other opcodes. If not enough resources can be found to
52
  allocate the instance, the temporary error code
53
  :pyeval:`errors.ECODE_TEMP_NORES` is returned. The operation can be
54
  retried thereafter, with or without opportunistic locking.
55
- The functionality for allocating multiple instances at once has been
56
  overhauled and is now also available through :doc:`RAPI <rapi>`.
57
- Man pages can now be included when the documentation is built, in
58
  which case the output is in ``doc/man-html``. The configure-time
59
  option is ``--enable-manpages-in-doc``. Sphinx 1.0 or higher is
60
  required.
61
- A new htool called ``hroller`` helps scheduling parallel shutdown of
62
  nodes on clusters where all instances are shut down, but without ever
63
  rebooting both primary and secondary node for an instance at the same
64
  time.
65
- Queries not needing locks or RPC access to the node can now be
66
  performed by the confd daemon, making them independent from jobs, and
67
  thus faster to execute. This is selectable at configure time.
68
- The KVM hypervisor has been updated with many new hypervisor
69
  parameters, including a generic one for passing arbitrary command line
70
  values. See a complete list in :manpage:`gnt-instance(8)`.
71
- A new tool, called ``mon-collector``, is the stand-alone executor of
72
  the data collectors for a monitoring system. As of this version, it
73
  just includes the DRBD data collector, that can be executed by calling
74
  ``mon-collector`` using the ``drbd`` parameter. See
75
  :manpage:`mon-collector(7)`.
76
- New ``exclusive-storage`` node parameter added. When it's set to true,
77
  physical disks are assigned in an exclusive fashion to instances, as
78
  documented in :doc:`Partitioned Ganeti <design-partitioned>`.
79
  Currently, only instances using the ``plain`` disk template are
80
  supported.
81
- New network management functionality to support automatic allocation
82
  of IP addresses and managing of network parameters. See
83
  :manpage:`gnt-network(8)` for more details.
84
- New external storage backend, to allow managing arbitrary storage
85
  systems external to the cluster. See
86
  :manpage:`ganeti-extstorage-interface(7)`.
87
- Instance renames of LVM-based instances will now update the LV tags
88
  (which can be used to recover the instance-to-LV mapping in case of
89
  emergencies)
90
- ``hbal`` will now exit with status 0 if, during job execution over
91
  LUXI, early exit has been requested and all jobs are successful;
92
  before, exit status 1 was used, which cannot be differentiated from
93
  "job error" case
94

    
95

    
96
Version 2.6.2
97
-------------
98

    
99
*(Released Fri, 21 Dec 2012)*
100

    
101
Important behaviour change: hbal won't rebalance anymore instances which
102
have the ``auto_balance`` attribute set to false. This was the intention
103
all along, but until now it only skipped those from the N+1 memory
104
reservation (DRBD-specific).
105

    
106
A significant number of bug fixes in this release:
107

    
108
- Fixed disk adoption interaction with ipolicy checks.
109
- Fixed networking issues when instances are started, stopped or
110
  migrated, by forcing the tap device's MAC prefix to "fe" (issue 217).
111
- Fixed the warning in cluster verify for shared storage instances not
112
  being redundant.
113
- Fixed removal of storage directory on shared file storage (issue 262).
114
- Fixed validation of LVM volume group name in OpClusterSetParams
115
  (``gnt-cluster modify``) (issue 285).
116
- Fixed runtime memory increases (``gnt-instance modify -m``).
117
- Fixed live migration under Xen's ``xl`` mode.
118
- Fixed ``gnt-instance console`` with ``xl``.
119
- Fixed building with newer Haskell compiler/libraries.
120
- Fixed PID file writing in Haskell daemons (confd); this prevents
121
  restart issues if confd was launched manually (outside of
122
  ``daemon-util``) while another copy of it was running
123
- Fixed a type error when doing live migrations with KVM (issue 297) and
124
  the error messages for failing migrations have been improved.
125
- Fixed opcode validation for the out-of-band commands (``gnt-node
126
  power``).
127
- Fixed a type error when unsetting OS hypervisor parameters (issue
128
  311); now it's possible to unset all OS-specific hypervisor
129
  parameters.
130
- Fixed the ``dry-run`` mode for many operations: verification of
131
  results was over-zealous but didn't take into account the ``dry-run``
132
  operation, resulting in "wrong" failures.
133
- Fixed bash completion in ``gnt-job list`` when the job queue has
134
  hundreds of entries; especially with older ``bash`` versions, this
135
  results in significant CPU usage.
136

    
137
And lastly, a few other improvements have been made:
138

    
139
- Added option to force master-failover without voting (issue 282).
140
- Clarified error message on lock conflict (issue 287).
141
- Logging of newly submitted jobs has been improved (issue 290).
142
- Hostname checks have been made uniform between instance rename and
143
  create (issue 291).
144
- The ``--submit`` option is now supported by ``gnt-debug delay``.
145
- Shutting down the master daemon by sending SIGTERM now stops it from
146
  processing jobs waiting for locks; instead, those jobs will be started
147
  once again after the master daemon is started the next time (issue
148
  296).
149
- Support for Xen's ``xl`` program has been improved (besides the fixes
150
  above).
151
- Reduced logging noise in the Haskell confd daemon (only show one log
152
  entry for each config reload, instead of two).
153
- Several man page updates and typo fixes.
154

    
155

    
156
Version 2.6.1
157
-------------
158

    
159
*(Released Fri, 12 Oct 2012)*
160

    
161
A small bugfix release. Among the bugs fixed:
162

    
163
- Fixed double use of ``PRIORITY_OPT`` in ``gnt-node migrate``, that
164
  made the command unusable.
165
- Commands that issue many jobs don't fail anymore just because some jobs
166
  take so long that other jobs are archived.
167
- Failures during ``gnt-instance reinstall`` are reflected by the exit
168
  status.
169
- Issue 190 fixed. Check for DRBD in cluster verify is enabled only when
170
  DRBD is enabled.
171
- When ``always_failover`` is set, ``--allow-failover`` is not required
172
  in migrate commands anymore.
173
- ``bash_completion`` works even if extglob is disabled.
174
- Fixed bug with locks that made failover for RDB-based instances fail.
175
- Fixed bug in non-mirrored instance allocation that made Ganeti choose
176
  a random node instead of one based on the allocator metric.
177
- Support for newer versions of pylint and pep8.
178
- Hail doesn't fail anymore when trying to add an instance of type
179
  ``file``, ``sharedfile`` or ``rbd``.
180
- Added new Makefile target to rebuild the whole distribution, so that
181
  all files are included.
182

    
183

    
184
Version 2.6.0
185
-------------
186

    
187
*(Released Fri, 27 Jul 2012)*
188

    
189

    
190
.. attention:: The ``LUXI`` protocol has been made more consistent
191
   regarding its handling of command arguments. This, however, leads to
192
   incompatibility issues with previous versions. Please ensure that you
193
   restart Ganeti daemons soon after the upgrade, otherwise most
194
   ``LUXI`` calls (job submission, setting/resetting the drain flag,
195
   pausing/resuming the watcher, cancelling and archiving jobs, querying
196
   the cluster configuration) will fail.
197

    
198

    
199
New features
200
~~~~~~~~~~~~
201

    
202
Instance run status
203
+++++++++++++++++++
204

    
205
The current ``admin_up`` field, which used to denote whether an instance
206
should be running or not, has been removed. Instead, ``admin_state`` is
207
introduced, with 3 possible values -- ``up``, ``down`` and ``offline``.
208

    
209
The rational behind this is that an instance being “down” can have
210
different meanings:
211

    
212
- it could be down during a reboot
213
- it could be temporarily be down for a reinstall
214
- or it could be down because it is deprecated and kept just for its
215
  disk
216

    
217
The previous Boolean state was making it difficult to do capacity
218
calculations: should Ganeti reserve memory for a down instance? Now, the
219
tri-state field makes it clear:
220

    
221
- in ``up`` and ``down`` state, all resources are reserved for the
222
  instance, and it can be at any time brought up if it is down
223
- in ``offline`` state, only disk space is reserved for it, but not
224
  memory or CPUs
225

    
226
The field can have an extra use: since the transition between ``up`` and
227
``down`` and vice-versus is done via ``gnt-instance start/stop``, but
228
transition between ``offline`` and ``down`` is done via ``gnt-instance
229
modify``, it is possible to given different rights to users. For
230
example, owners of an instance could be allowed to start/stop it, but
231
not transition it out of the offline state.
232

    
233
Instance policies and specs
234
+++++++++++++++++++++++++++
235

    
236
In previous Ganeti versions, an instance creation request was not
237
limited on the minimum size and on the maximum size just by the cluster
238
resources. As such, any policy could be implemented only in third-party
239
clients (RAPI clients, or shell wrappers over ``gnt-*``
240
tools). Furthermore, calculating cluster capacity via ``hspace`` again
241
required external input with regards to instance sizes.
242

    
243
In order to improve these workflows and to allow for example better
244
per-node group differentiation, we introduced instance specs, which
245
allow declaring:
246

    
247
- minimum instance disk size, disk count, memory size, cpu count
248
- maximum values for the above metrics
249
- and “standard” values (used in ``hspace`` to calculate the standard
250
  sized instances)
251

    
252
The minimum/maximum values can be also customised at node-group level,
253
for example allowing more powerful hardware to support bigger instance
254
memory sizes.
255

    
256
Beside the instance specs, there are a few other settings belonging to
257
the instance policy framework. It is possible now to customise, per
258
cluster and node-group:
259

    
260
- the list of allowed disk templates
261
- the maximum ratio of VCPUs per PCPUs (to control CPU oversubscription)
262
- the maximum ratio of instance to spindles (see below for more
263
  information) for local storage
264

    
265
All these together should allow all tools that talk to Ganeti to know
266
what are the ranges of allowed values for instances and the
267
over-subscription that is allowed.
268

    
269
For the VCPU/PCPU ratio, we already have the VCPU configuration from the
270
instance configuration, and the physical CPU configuration from the
271
node. For the spindle ratios however, we didn't track before these
272
values, so new parameters have been added:
273

    
274
- a new node parameter ``spindle_count``, defaults to 1, customisable at
275
  node group or node level
276
- at new backend parameter (for instances), ``spindle_use`` defaults to 1
277

    
278
Note that spindles in this context doesn't need to mean actual
279
mechanical hard-drives; it's just a relative number for both the node
280
I/O capacity and instance I/O consumption.
281

    
282
Instance migration behaviour
283
++++++++++++++++++++++++++++
284

    
285
While live-migration is in general desirable over failover, it is
286
possible that for some workloads it is actually worse, due to the
287
variable time of the “suspend” phase during live migration.
288

    
289
To allow the tools to work consistently over such instances (without
290
having to hard-code instance names), a new backend parameter
291
``always_failover`` has been added to control the migration/failover
292
behaviour. When set to True, all migration requests for an instance will
293
instead fall-back to failover.
294

    
295
Instance memory ballooning
296
++++++++++++++++++++++++++
297

    
298
Initial support for memory ballooning has been added. The memory for an
299
instance is no longer fixed (backend parameter ``memory``), but instead
300
can vary between minimum and maximum values (backend parameters
301
``minmem`` and ``maxmem``). Currently we only change an instance's
302
memory when:
303

    
304
- live migrating or failing over and instance and the target node
305
  doesn't have enough memory
306
- user requests changing the memory via ``gnt-instance modify
307
  --runtime-memory``
308

    
309
Instance CPU pinning
310
++++++++++++++++++++
311

    
312
In order to control the use of specific CPUs by instance, support for
313
controlling CPU pinning has been added for the Xen, HVM and LXC
314
hypervisors. This is controlled by a new hypervisor parameter
315
``cpu_mask``; details about possible values for this are in the
316
:manpage:`gnt-instance(8)`. Note that use of the most specific (precise
317
VCPU-to-CPU mapping) form will work well only when all nodes in your
318
cluster have the same amount of CPUs.
319

    
320
Disk parameters
321
+++++++++++++++
322

    
323
Another area in which Ganeti was not customisable were the parameters
324
used for storage configuration, e.g. how many stripes to use for LVM,
325
DRBD resync configuration, etc.
326

    
327
To improve this area, we've added disks parameters, which are
328
customisable at cluster and node group level, and which allow to
329
specify various parameters for disks (DRBD has the most parameters
330
currently), for example:
331

    
332
- DRBD resync algorithm and parameters (e.g. speed)
333
- the default VG for meta-data volumes for DRBD
334
- number of stripes for LVM (plain disk template)
335
- the RBD pool
336

    
337
These parameters can be modified via ``gnt-cluster modify -D …`` and
338
``gnt-group modify -D …``, and are used at either instance creation (in
339
case of LVM stripes, for example) or at disk “activation” time
340
(e.g. resync speed).
341

    
342
Rados block device support
343
++++++++++++++++++++++++++
344

    
345
A Rados (http://ceph.com/wiki/Rbd) storage backend has been added,
346
denoted by the ``rbd`` disk template type. This is considered
347
experimental, feedback is welcome. For details on configuring it, see
348
the :doc:`install` document and the :manpage:`gnt-cluster(8)` man page.
349

    
350
Master IP setup
351
+++++++++++++++
352

    
353
The existing master IP functionality works well only in simple setups (a
354
single network shared by all nodes); however, if nodes belong to
355
different networks, then the ``/32`` setup and lack of routing
356
information is not enough.
357

    
358
To allow the master IP to function well in more complex cases, the
359
system was reworked as follows:
360

    
361
- a master IP netmask setting has been added
362
- the master IP activation/turn-down code was moved from the node daemon
363
  to a separate script
364
- whether to run the Ganeti-supplied master IP script or a user-supplied
365
  on is a ``gnt-cluster init`` setting
366

    
367
Details about the location of the standard and custom setup scripts are
368
in the man page :manpage:`gnt-cluster(8)`; for information about the
369
setup script protocol, look at the Ganeti-supplied script.
370

    
371
SPICE support
372
+++++++++++++
373

    
374
The `SPICE <http://www.linux-kvm.org/page/SPICE>`_ support has been
375
improved.
376

    
377
It is now possible to use TLS-protected connections, and when renewing
378
or changing the cluster certificates (via ``gnt-cluster renew-crypto``,
379
it is now possible to specify spice or spice CA certificates. Also, it
380
is possible to configure a password for SPICE sessions via the
381
hypervisor parameter ``spice_password_file``.
382

    
383
There are also new parameters to control the compression and streaming
384
options (e.g. ``spice_image_compression``, ``spice_streaming_video``,
385
etc.). For details, see the man page :manpage:`gnt-instance(8)` and look
386
for the spice parameters.
387

    
388
Lastly, it is now possible to see the SPICE connection information via
389
``gnt-instance console``.
390

    
391
OVF converter
392
+++++++++++++
393

    
394
A new tool (``tools/ovfconverter``) has been added that supports
395
conversion between Ganeti and the `Open Virtualization Format
396
<http://en.wikipedia.org/wiki/Open_Virtualization_Format>`_ (both to and
397
from).
398

    
399
This relies on the ``qemu-img`` tool to convert the disk formats, so the
400
actual compatibility with other virtualization solutions depends on it.
401

    
402
Confd daemon changes
403
++++++++++++++++++++
404

    
405
The configuration query daemon (``ganeti-confd``) is now optional, and
406
has been rewritten in Haskell; whether to use the daemon at all, use the
407
Python (default) or the Haskell version is selectable at configure time
408
via the ``--enable-confd`` parameter, which can take one of the
409
``haskell``, ``python`` or ``no`` values. If not used, disabling the
410
daemon will result in a smaller footprint; for larger systems, we
411
welcome feedback on the Haskell version which might become the default
412
in future versions.
413

    
414
If you want to use ``gnt-node list-drbd`` you need to have the Haskell
415
daemon running. The Python version doesn't implement the new call.
416

    
417

    
418
User interface changes
419
~~~~~~~~~~~~~~~~~~~~~~
420

    
421
We have replaced the ``--disks`` option of ``gnt-instance
422
replace-disks`` with a more flexible ``--disk`` option, which allows
423
adding and removing disks at arbitrary indices (Issue 188). Furthermore,
424
disk size and mode can be changed upon recreation (via ``gnt-instance
425
recreate-disks``, which accepts the same ``--disk`` option).
426

    
427
As many people are used to a ``show`` command, we have added that as an
428
alias to ``info`` on all ``gnt-*`` commands.
429

    
430
The ``gnt-instance grow-disk`` command has a new mode in which it can
431
accept the target size of the disk, instead of the delta; this can be
432
more safe since two runs in absolute mode will be idempotent, and
433
sometimes it's also easier to specify the desired size directly.
434

    
435
Also the handling of instances with regard to offline secondaries has
436
been improved. Instance operations should not fail because one of it's
437
secondary nodes is offline, even though it's safe to proceed.
438

    
439
A new command ``list-drbd`` has been added to the ``gnt-node`` script to
440
support debugging of DRBD issues on nodes. It provides a mapping of DRBD
441
minors to instance name.
442

    
443
API changes
444
~~~~~~~~~~~
445

    
446
RAPI coverage has improved, with (for example) new resources for
447
recreate-disks, node power-cycle, etc.
448

    
449
Compatibility
450
~~~~~~~~~~~~~
451

    
452
There is partial support for ``xl`` in the Xen hypervisor; feedback is
453
welcome.
454

    
455
Python 2.7 is better supported, and after Ganeti 2.6 we will investigate
456
whether to still support Python 2.4 or move to Python 2.6 as minimum
457
required version.
458

    
459
Support for Fedora has been slightly improved; the provided example
460
init.d script should work better on it and the INSTALL file should
461
document the needed dependencies.
462

    
463
Internal changes
464
~~~~~~~~~~~~~~~~
465

    
466
The deprecated ``QueryLocks`` LUXI request has been removed. Use
467
``Query(what=QR_LOCK, ...)`` instead.
468

    
469
The LUXI requests :pyeval:`luxi.REQ_QUERY_JOBS`,
470
:pyeval:`luxi.REQ_QUERY_INSTANCES`, :pyeval:`luxi.REQ_QUERY_NODES`,
471
:pyeval:`luxi.REQ_QUERY_GROUPS`, :pyeval:`luxi.REQ_QUERY_EXPORTS` and
472
:pyeval:`luxi.REQ_QUERY_TAGS` are deprecated and will be removed in a
473
future version. :pyeval:`luxi.REQ_QUERY` should be used instead.
474

    
475
RAPI client: ``CertificateError`` now derives from
476
``GanetiApiError``. This should make it more easy to handle Ganeti
477
errors.
478

    
479
Deprecation warnings due to PyCrypto/paramiko import in
480
``tools/setup-ssh`` have been silenced, as usually they are safe; please
481
make sure to run an up-to-date paramiko version, if you use this tool.
482

    
483
The QA scripts now depend on Python 2.5 or above (the main code base
484
still works with Python 2.4).
485

    
486
The configuration file (``config.data``) is now written without
487
indentation for performance reasons; if you want to edit it, it can be
488
re-formatted via ``tools/fmtjson``.
489

    
490
A number of bugs has been fixed in the cluster merge tool.
491

    
492
``x509`` certification verification (used in import-export) has been
493
changed to allow the same clock skew as permitted by the cluster
494
verification. This will remove some rare but hard to diagnose errors in
495
import-export.
496

    
497

    
498
Version 2.6.0 rc4
499
-----------------
500

    
501
*(Released Thu, 19 Jul 2012)*
502

    
503
Very few changes from rc4 to the final release, only bugfixes:
504

    
505
- integrated fixes from release 2.5.2 (fix general boot flag for KVM
506
  instance, fix CDROM booting for KVM instances)
507
- fixed node group modification of node parameters
508
- fixed issue in LUClusterVerifyGroup with multi-group clusters
509
- fixed generation of bash completion to ensure a stable ordering
510
- fixed a few typos
511

    
512

    
513
Version 2.6.0 rc3
514
-----------------
515

    
516
*(Released Fri, 13 Jul 2012)*
517

    
518
Third release candidate for 2.6. The following changes were done from
519
rc3 to rc4:
520

    
521
- Fixed ``UpgradeConfig`` w.r.t. to disk parameters on disk objects.
522
- Fixed an inconsistency in the LUXI protocol with the provided
523
  arguments (NOT backwards compatible)
524
- Fixed a bug with node groups ipolicy where ``min`` was greater than
525
  the cluster ``std`` value
526
- Implemented a new ``gnt-node list-drbd`` call to list DRBD minors for
527
  easier instance debugging on nodes (requires ``hconfd`` to work)
528

    
529

    
530
Version 2.6.0 rc2
531
-----------------
532

    
533
*(Released Tue, 03 Jul 2012)*
534

    
535
Second release candidate for 2.6. The following changes were done from
536
rc2 to rc3:
537

    
538
- Fixed ``gnt-cluster verify`` regarding ``master-ip-script`` on non
539
  master candidates
540
- Fixed a RAPI regression on missing beparams/memory
541
- Fixed redistribution of files on offline nodes
542
- Added possibility to run activate-disks even though secondaries are
543
  offline. With this change it relaxes also the strictness on some other
544
  commands which use activate disks internally:
545
  * ``gnt-instance start|reboot|rename|backup|export``
546
- Made it possible to remove safely an instance if its secondaries are
547
  offline
548
- Made it possible to reinstall even though secondaries are offline
549

    
550

    
551
Version 2.6.0 rc1
552
-----------------
553

    
554
*(Released Mon, 25 Jun 2012)*
555

    
556
First release candidate for 2.6. The following changes were done from
557
rc1 to rc2:
558

    
559
- Fixed bugs with disk parameters and ``rbd`` templates as well as
560
  ``instance_os_add``
561
- Made ``gnt-instance modify`` more consistent regarding new NIC/Disk
562
  behaviour. It supports now the modify operation
563
- ``hcheck`` implemented to analyze cluster health and possibility of
564
  improving health by rebalance
565
- ``hbal`` has been improved in dealing with split instances
566

    
567

    
568
Version 2.6.0 beta2
569
-------------------
570

    
571
*(Released Mon, 11 Jun 2012)*
572

    
573
Second beta release of 2.6. The following changes were done from beta2
574
to rc1:
575

    
576
- Fixed ``daemon-util`` with non-root user models
577
- Fixed creation of plain instances with ``--no-wait-for-sync``
578
- Fix wrong iv_names when running ``cfgupgrade``
579
- Export more information in RAPI group queries
580
- Fixed bug when changing instance network interfaces
581
- Extended burnin to do NIC changes
582
- query: Added ``<``, ``>``, ``<=``, ``>=`` comparison operators
583
- Changed default for DRBD barriers
584
- Fixed DRBD error reporting for syncer rate
585
- Verify the options on disk parameters
586

    
587
And of course various fixes to documentation and improved unittests and
588
QA.
589

    
590

    
591
Version 2.6.0 beta1
592
-------------------
593

    
594
*(Released Wed, 23 May 2012)*
595

    
596
First beta release of 2.6. The following changes were done from beta1 to
597
beta2:
598

    
599
- integrated patch for distributions without ``start-stop-daemon``
600
- adapted example init.d script to work on Fedora
601
- fixed log handling in Haskell daemons
602
- adapted checks in the watcher for pycurl linked against libnss
603
- add partial support for ``xl`` instead of ``xm`` for Xen
604
- fixed a type issue in cluster verification
605
- fixed ssconf handling in the Haskell code (was breaking confd in IPv6
606
  clusters)
607

    
608
Plus integrated fixes from the 2.5 branch:
609

    
610
- fixed ``kvm-ifup`` to use ``/bin/bash``
611
- fixed parallel build failures
612
- KVM live migration when using a custom keymap
613

    
614

    
615
Version 2.5.2
616
-------------
617

    
618
*(Released Tue, 24 Jul 2012)*
619

    
620
A small bugfix release, with no new features:
621

    
622
- fixed bash-isms in kvm-ifup, for compatibility with systems which use a
623
  different default shell (e.g. Debian, Ubuntu)
624
- fixed KVM startup and live migration with a custom keymap (fixes Issue
625
  243 and Debian bug #650664)
626
- fixed compatibility with KVM versions that don't support multiple boot
627
  devices (fixes Issue 230 and Debian bug #624256)
628

    
629
Additionally, a few fixes were done to the build system (fixed parallel
630
build failures) and to the unittests (fixed race condition in test for
631
FileID functions, and the default enable/disable mode for QA test is now
632
customisable).
633

    
634

    
635
Version 2.5.1
636
-------------
637

    
638
*(Released Fri, 11 May 2012)*
639

    
640
A small bugfix release.
641

    
642
The main issues solved are on the topic of compatibility with newer LVM
643
releases:
644

    
645
- fixed parsing of ``lv_attr`` field
646
- adapted to new ``vgreduce --removemissing`` behaviour where sometimes
647
  the ``--force`` flag is needed
648

    
649
Also on the topic of compatibility, ``tools/lvmstrap`` has been changed
650
to accept kernel 3.x too (was hardcoded to 2.6.*).
651

    
652
A regression present in 2.5.0 that broke handling (in the gnt-* scripts)
653
of hook results and that also made display of other errors suboptimal
654
was fixed; the code behaves now like 2.4 and earlier.
655

    
656
Another change in 2.5, the cleanup of the OS scripts environment, is too
657
aggressive: it removed even the ``PATH`` variable, which requires the OS
658
scripts to *always* need to export it. Since this is a bit too strict,
659
we now export a minimal PATH, the same that we export for hooks.
660

    
661
The fix for issue 201 (Preserve bridge MTU in KVM ifup script) was
662
integrated into this release.
663

    
664
Finally, a few other miscellaneous changes were done (no new features,
665
just small improvements):
666

    
667
- Fix ``gnt-group --help`` display
668
- Fix hardcoded Xen kernel path
669
- Fix grow-disk handling of invalid units
670
- Update synopsis for ``gnt-cluster repair-disk-sizes``
671
- Accept both PUT and POST in noded (makes future upgrade to 2.6 easier)
672

    
673

    
674
Version 2.5.0
675
-------------
676

    
677
*(Released Thu, 12 Apr 2012)*
678

    
679
Incompatible/important changes and bugfixes
680
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
681

    
682
- The default of the ``/2/instances/[instance_name]/rename`` RAPI
683
  resource's ``ip_check`` parameter changed from ``True`` to ``False``
684
  to match the underlying LUXI interface.
685
- The ``/2/nodes/[node_name]/evacuate`` RAPI resource was changed to use
686
  body parameters, see :doc:`RAPI documentation <rapi>`. The server does
687
  not maintain backwards-compatibility as the underlying operation
688
  changed in an incompatible way. The RAPI client can talk to old
689
  servers, but it needs to be told so as the return value changed.
690
- When creating file-based instances via RAPI, the ``file_driver``
691
  parameter no longer defaults to ``loop`` and must be specified.
692
- The deprecated ``bridge`` NIC parameter is no longer supported. Use
693
  ``link`` instead.
694
- Support for the undocumented and deprecated RAPI instance creation
695
  request format version 0 has been dropped. Use version 1, supported
696
  since Ganeti 2.1.3 and :doc:`documented <rapi>`, instead.
697
- Pyparsing 1.4.6 or above is required, see :doc:`installation
698
  documentation <install>`.
699
- The "cluster-verify" hooks are now executed per group by the
700
  ``OP_CLUSTER_VERIFY_GROUP`` opcode. This maintains the same behavior
701
  if you just run ``gnt-cluster verify``, which generates one opcode per
702
  group.
703
- The environment as passed to the OS scripts is cleared, and thus no
704
  environment variables defined in the node daemon's environment will be
705
  inherited by the scripts.
706
- The :doc:`iallocator <iallocator>` mode ``multi-evacuate`` has been
707
  deprecated.
708
- :doc:`New iallocator modes <design-multi-reloc>` have been added to
709
  support operations involving multiple node groups.
710
- Offline nodes are ignored when failing over an instance.
711
- Support for KVM version 1.0, which changed the version reporting format
712
  from 3 to 2 digits.
713
- TCP/IP ports used by DRBD disks are returned to a pool upon instance
714
  removal.
715
- ``Makefile`` is now compatible with Automake 1.11.2
716
- Includes all bugfixes made in the 2.4 series
717

    
718
New features
719
~~~~~~~~~~~~
720

    
721
- The ganeti-htools project has been merged into the ganeti-core source
722
  tree and will be built as part of Ganeti (see :doc:`install-quick`).
723
- Implemented support for :doc:`shared storage <design-shared-storage>`.
724
- Add support for disks larger than 2 TB in ``lvmstrap`` by supporting
725
  GPT-style partition tables (requires `parted
726
  <http://www.gnu.org/s/parted/>`_).
727
- Added support for floppy drive and 2nd CD-ROM drive in KVM hypervisor.
728
- Allowed adding tags on instance creation.
729
- Export instance tags to hooks (``INSTANCE_TAGS``, see :doc:`hooks`)
730
- Allow instances to be started in a paused state, enabling the user to
731
  see the complete console output on boot using the console.
732
- Added new hypervisor flag to control default reboot behaviour
733
  (``reboot_behavior``).
734
- Added support for KVM keymaps (hypervisor parameter ``keymap``).
735
- Improved out-of-band management support:
736

    
737
  - Added ``gnt-node health`` command reporting the health status of
738
    nodes.
739
  - Added ``gnt-node power`` command to manage power status of nodes.
740
  - Added command for emergency power-off (EPO), ``gnt-cluster epo``.
741

    
742
- Instance migration can fall back to failover if instance is not
743
  running.
744
- Filters can be used when listing nodes, instances, groups and locks;
745
  see :manpage:`ganeti(7)` manpage.
746
- Added post-execution status as variables to :doc:`hooks <hooks>`
747
  environment.
748
- Instance tags are exported/imported together with the instance.
749
- When given an explicit job ID, ``gnt-job info`` will work for archived
750
  jobs.
751
- Jobs can define dependencies on other jobs (not yet supported via
752
  RAPI or command line, but used by internal commands and usable via
753
  LUXI).
754

    
755
  - Lock monitor (``gnt-debug locks``) shows jobs waiting for
756
    dependencies.
757

    
758
- Instance failover is now available as a RAPI resource
759
  (``/2/instances/[instance_name]/failover``).
760
- ``gnt-instance info`` defaults to static information if primary node
761
  is offline.
762
- Opcodes have a new ``comment`` attribute.
763
- Added basic SPICE support to KVM hypervisor.
764
- ``tools/ganeti-listrunner`` allows passing of arguments to executable.
765

    
766
Node group improvements
767
~~~~~~~~~~~~~~~~~~~~~~~
768

    
769
- ``gnt-cluster verify`` has been modified to check groups separately,
770
  thereby improving performance.
771
- Node group support has been added to ``gnt-cluster verify-disks``,
772
  which now operates per node group.
773
- Watcher has been changed to work better with node groups.
774

    
775
  - One process and state file per node group.
776
  - Slow watcher in one group doesn't block other group's watcher.
777

    
778
- Added new command, ``gnt-group evacuate``, to move all instances in a
779
  node group to other groups.
780
- Added ``gnt-instance change-group`` to move an instance to another
781
  node group.
782
- ``gnt-cluster command`` and ``gnt-cluster copyfile`` now support
783
  per-group operations.
784
- Node groups can be tagged.
785
- Some operations switch from an exclusive to a shared lock as soon as
786
  possible.
787
- Instance's primary and secondary nodes' groups are now available as
788
  query fields (``pnode.group``, ``pnode.group.uuid``, ``snodes.group``
789
  and ``snodes.group.uuid``).
790

    
791
Misc
792
~~~~
793

    
794
- Numerous updates to documentation and manpages.
795

    
796
  - :doc:`RAPI <rapi>` documentation now has detailed parameter
797
    descriptions.
798
  - Some opcode/job results are now also documented, see :doc:`RAPI
799
    <rapi>`.
800

    
801
- A lockset's internal lock is now also visible in lock monitor.
802
- Log messages from job queue workers now contain information about the
803
  opcode they're processing.
804
- ``gnt-instance console`` no longer requires the instance lock.
805
- A short delay when waiting for job changes reduces the number of LUXI
806
  requests significantly.
807
- DRBD metadata volumes are overwritten with zeros during disk creation.
808
- Out-of-band commands no longer acquire the cluster lock in exclusive
809
  mode.
810
- ``devel/upload`` now uses correct permissions for directories.
811

    
812

    
813
Version 2.5.0 rc6
814
-----------------
815

    
816
*(Released Fri, 23 Mar 2012)*
817

    
818
This was the sixth release candidate of the 2.5 series.
819

    
820

    
821
Version 2.5.0 rc5
822
-----------------
823

    
824
*(Released Mon, 9 Jan 2012)*
825

    
826
This was the fifth release candidate of the 2.5 series.
827

    
828

    
829
Version 2.5.0 rc4
830
-----------------
831

    
832
*(Released Thu, 27 Oct 2011)*
833

    
834
This was the fourth release candidate of the 2.5 series.
835

    
836

    
837
Version 2.5.0 rc3
838
-----------------
839

    
840
*(Released Wed, 26 Oct 2011)*
841

    
842
This was the third release candidate of the 2.5 series.
843

    
844

    
845
Version 2.5.0 rc2
846
-----------------
847

    
848
*(Released Tue, 18 Oct 2011)*
849

    
850
This was the second release candidate of the 2.5 series.
851

    
852

    
853
Version 2.5.0 rc1
854
-----------------
855

    
856
*(Released Tue, 4 Oct 2011)*
857

    
858
This was the first release candidate of the 2.5 series.
859

    
860

    
861
Version 2.5.0 beta3
862
-------------------
863

    
864
*(Released Wed, 31 Aug 2011)*
865

    
866
This was the third beta release of the 2.5 series.
867

    
868

    
869
Version 2.5.0 beta2
870
-------------------
871

    
872
*(Released Mon, 22 Aug 2011)*
873

    
874
This was the second beta release of the 2.5 series.
875

    
876

    
877
Version 2.5.0 beta1
878
-------------------
879

    
880
*(Released Fri, 12 Aug 2011)*
881

    
882
This was the first beta release of the 2.5 series.
883

    
884

    
885
Version 2.4.5
886
-------------
887

    
888
*(Released Thu, 27 Oct 2011)*
889

    
890
- Fixed bug when parsing command line parameter values ending in
891
  backslash
892
- Fixed assertion error after unclean master shutdown
893
- Disable HTTP client pool for RPC, significantly reducing memory usage
894
  of master daemon
895
- Fixed queue archive creation with wrong permissions
896

    
897

    
898
Version 2.4.4
899
-------------
900

    
901
*(Released Tue, 23 Aug 2011)*
902

    
903
Small bug-fixes:
904

    
905
- Fixed documentation for importing with ``--src-dir`` option
906
- Fixed a bug in ``ensure-dirs`` with queue/archive permissions
907
- Fixed a parsing issue with DRBD 8.3.11 in the Linux kernel
908

    
909

    
910
Version 2.4.3
911
-------------
912

    
913
*(Released Fri, 5 Aug 2011)*
914

    
915
Many bug-fixes and a few small features:
916

    
917
- Fixed argument order in ``ReserveLV`` and ``ReserveMAC`` which caused
918
  issues when you tried to add an instance with two MAC addresses in one
919
  request
920
- KVM: fixed per-instance stored UID value
921
- KVM: configure bridged NICs at migration start
922
- KVM: Fix a bug where instance will not start with never KVM versions
923
  (>= 0.14)
924
- Added OS search path to ``gnt-cluster info``
925
- Fixed an issue with ``file_storage_dir`` where you were forced to
926
  provide an absolute path, but the documentation states it is a
927
  relative path, the documentation was right
928
- Added a new parameter to instance stop/start called ``--no-remember``
929
  that will make the state change to not be remembered
930
- Implemented ``no_remember`` at RAPI level
931
- Improved the documentation
932
- Node evacuation: don't call IAllocator if node is already empty
933
- Fixed bug in DRBD8 replace disks on current nodes
934
- Fixed bug in recreate-disks for DRBD instances
935
- Moved assertion checking locks in ``gnt-instance replace-disks``
936
  causing it to abort with not owning the right locks for some situation
937
- Job queue: Fixed potential race condition when cancelling queued jobs
938
- Fixed off-by-one bug in job serial generation
939
- ``gnt-node volumes``: Fix instance names
940
- Fixed aliases in bash completion
941
- Fixed a bug in reopening log files after being sent a SIGHUP
942
- Added a flag to burnin to allow specifying VCPU count
943
- Bugfixes to non-root Ganeti configuration
944

    
945

    
946
Version 2.4.2
947
-------------
948

    
949
*(Released Thu, 12 May 2011)*
950

    
951
Many bug-fixes and a few new small features:
952

    
953
- Fixed a bug related to log opening failures
954
- Fixed a bug in instance listing with orphan instances
955
- Fixed a bug which prevented resetting the cluster-level node parameter
956
  ``oob_program`` to the default
957
- Many fixes related to the ``cluster-merge`` tool
958
- Fixed a race condition in the lock monitor, which caused failures
959
  during (at least) creation of many instances in parallel
960
- Improved output for gnt-job info
961
- Removed the quiet flag on some ssh calls which prevented debugging
962
  failures
963
- Improved the N+1 failure messages in cluster verify by actually
964
  showing the memory values (needed and available)
965
- Increased lock attempt timeouts so that when executing long operations
966
  (e.g. DRBD replace-disks) other jobs do not enter 'blocking acquire'
967
  too early and thus prevent the use of the 'fair' mechanism
968
- Changed instance query data (``gnt-instance info``) to not acquire
969
  locks unless needed, thus allowing its use on locked instance if only
970
  static information is asked for
971
- Improved behaviour with filesystems that do not support rename on an
972
  opened file
973
- Fixed the behaviour of ``prealloc_wipe_disks`` cluster parameter which
974
  kept locks on all nodes during the wipe, which is unneeded
975
- Fixed ``gnt-watcher`` handling of errors during hooks execution
976
- Fixed bug in ``prealloc_wipe_disks`` with small disk sizes (less than
977
  10GiB) which caused the wipe to fail right at the end in some cases
978
- Fixed master IP activation when doing master failover with no-voting
979
- Fixed bug in ``gnt-node add --readd`` which allowed the re-adding of
980
  the master node itself
981
- Fixed potential data-loss in under disk full conditions, where Ganeti
982
  wouldn't check correctly the return code and would consider
983
  partially-written files 'correct'
984
- Fixed bug related to multiple VGs and DRBD disk replacing
985
- Added new disk parameter ``metavg`` that allows placement of the meta
986
  device for DRBD in a different volume group
987
- Fixed error handling in the node daemon when the system libc doesn't
988
  have major number 6 (i.e. if ``libc.so.6`` is not the actual libc)
989
- Fixed lock release during replace-disks, which kept cluster-wide locks
990
  when doing disk replaces with an iallocator script
991
- Added check for missing bridges in cluster verify
992
- Handle EPIPE errors while writing to the terminal better, so that
993
  piping the output to e.g. ``less`` doesn't cause a backtrace
994
- Fixed rare case where a ^C during Luxi calls could have been
995
  interpreted as server errors, instead of simply terminating
996
- Fixed a race condition in LUGroupAssignNodes (``gnt-group
997
  assign-nodes``)
998
- Added a few more parameters to the KVM hypervisor, allowing a second
999
  CDROM, custom disk type for CDROMs and a floppy image
1000
- Removed redundant message in instance rename when the name is given
1001
  already as a FQDN
1002
- Added option to ``gnt-instance recreate-disks`` to allow creating the
1003
  disks on new nodes, allowing recreation when the original instance
1004
  nodes are completely gone
1005
- Added option when converting disk templates to DRBD to skip waiting
1006
  for the resync, in order to make the instance available sooner
1007
- Added two new variables to the OS scripts environment (containing the
1008
  instance's nodes)
1009
- Made the root_path and optional parameter for the xen-pvm hypervisor,
1010
  to allow use of ``pvgrub`` as bootloader
1011
- Changed the instance memory modifications to only check out-of-memory
1012
  conditions on memory increases, and turned the secondary node warnings
1013
  into errors (they can still be overridden via ``--force``)
1014
- Fixed the handling of a corner case when the Python installation gets
1015
  corrupted (e.g. a bad disk) while ganeti-noded is running and we try
1016
  to execute a command that doesn't exist
1017
- Fixed a bug in ``gnt-instance move`` (LUInstanceMove) when the primary
1018
  node of the instance returned failures during instance shutdown; this
1019
  adds the option ``--ignore-consistency`` to gnt-instance move
1020

    
1021
And as usual, various improvements to the error messages, documentation
1022
and man pages.
1023

    
1024

    
1025
Version 2.4.1
1026
-------------
1027

    
1028
*(Released Wed, 09 Mar 2011)*
1029

    
1030
Emergency bug-fix release. ``tools/cfgupgrade`` was broken and overwrote
1031
the RAPI users file if run twice (even with ``--dry-run``).
1032

    
1033
The release fixes that bug (nothing else changed).
1034

    
1035

    
1036
Version 2.4.0
1037
-------------
1038

    
1039
*(Released Mon, 07 Mar 2011)*
1040

    
1041
Final 2.4.0 release. Just a few small fixes:
1042

    
1043
- Fixed RAPI node evacuate
1044
- Fixed the kvm-ifup script
1045
- Fixed internal error handling for special job cases
1046
- Updated man page to specify the escaping feature for options
1047

    
1048

    
1049
Version 2.4.0 rc3
1050
-----------------
1051

    
1052
*(Released Mon, 28 Feb 2011)*
1053

    
1054
A critical fix for the ``prealloc_wipe_disks`` feature: it is possible
1055
that this feature wiped the disks of the wrong instance, leading to loss
1056
of data.
1057

    
1058
Other changes:
1059

    
1060
- Fixed title of query field containing instance name
1061
- Expanded the glossary in the documentation
1062
- Fixed one unittest (internal issue)
1063

    
1064

    
1065
Version 2.4.0 rc2
1066
-----------------
1067

    
1068
*(Released Mon, 21 Feb 2011)*
1069

    
1070
A number of bug fixes plus just a couple functionality changes.
1071

    
1072
On the user-visible side, the ``gnt-* list`` command output has changed
1073
with respect to "special" field states. The current rc1 style of display
1074
can be re-enabled by passing a new ``--verbose`` (``-v``) flag, but in
1075
the default output mode special fields states are displayed as follows:
1076

    
1077
- Offline resource: ``*``
1078
- Unavailable/not applicable: ``-``
1079
- Data missing (RPC failure): ``?``
1080
- Unknown field: ``??``
1081

    
1082
Another user-visible change is the addition of ``--force-join`` to
1083
``gnt-node add``.
1084

    
1085
As for bug fixes:
1086

    
1087
- ``tools/cluster-merge`` has seen many fixes and is now enabled again
1088
- Fixed regression in RAPI/instance reinstall where all parameters were
1089
  required (instead of optional)
1090
- Fixed ``gnt-cluster repair-disk-sizes``, was broken since Ganeti 2.2
1091
- Fixed iallocator usage (offline nodes were not considered offline)
1092
- Fixed ``gnt-node list`` with respect to non-vm_capable nodes
1093
- Fixed hypervisor and OS parameter validation with respect to
1094
  non-vm_capable nodes
1095
- Fixed ``gnt-cluster verify`` with respect to offline nodes (mostly
1096
  cosmetic)
1097
- Fixed ``tools/listrunner`` with respect to agent-based usage
1098

    
1099

    
1100
Version 2.4.0 rc1
1101
-----------------
1102

    
1103
*(Released Fri,  4 Feb 2011)*
1104

    
1105
Many changes and fixes since the beta1 release. While there were some
1106
internal changes, the code has been mostly stabilised for the RC
1107
release.
1108

    
1109
Note: the dumb allocator was removed in this release, as it was not kept
1110
up-to-date with the IAllocator protocol changes. It is recommended to
1111
use the ``hail`` command from the ganeti-htools package.
1112

    
1113
Note: the 2.4 and up versions of Ganeti are not compatible with the
1114
0.2.x branch of ganeti-htools. You need to upgrade to
1115
ganeti-htools-0.3.0 (or later).
1116

    
1117
Regressions fixed from 2.3
1118
~~~~~~~~~~~~~~~~~~~~~~~~~~
1119

    
1120
- Fixed the ``gnt-cluster verify-disks`` command
1121
- Made ``gnt-cluster verify-disks`` work in parallel (as opposed to
1122
  serially on nodes)
1123
- Fixed disk adoption breakage
1124
- Fixed wrong headers in instance listing for field aliases
1125

    
1126
Other bugs fixed
1127
~~~~~~~~~~~~~~~~
1128

    
1129
- Fixed corner case in KVM handling of NICs
1130
- Fixed many cases of wrong handling of non-vm_capable nodes
1131
- Fixed a bug where a missing instance symlink was not possible to
1132
  recreate with any ``gnt-*`` command (now ``gnt-instance
1133
  activate-disks`` does it)
1134
- Fixed the volume group name as reported by ``gnt-cluster
1135
  verify-disks``
1136
- Increased timeouts for the import-export code, hopefully leading to
1137
  fewer aborts due network or instance timeouts
1138
- Fixed bug in ``gnt-node list-storage``
1139
- Fixed bug where not all daemons were started on cluster
1140
  initialisation, but only at the first watcher run
1141
- Fixed many bugs in the OOB implementation
1142
- Fixed watcher behaviour in presence of instances with offline
1143
  secondaries
1144
- Fixed instance list output for instances running on the wrong node
1145
- a few fixes to the cluster-merge tool, but it still cannot merge
1146
  multi-node groups (currently it is not recommended to use this tool)
1147

    
1148

    
1149
Improvements
1150
~~~~~~~~~~~~
1151

    
1152
- Improved network configuration for the KVM hypervisor
1153
- Added e1000 as a supported NIC for Xen-HVM
1154
- Improved the lvmstrap tool to also be able to use partitions, as
1155
  opposed to full disks
1156
- Improved speed of disk wiping (the cluster parameter
1157
  ``prealloc_wipe_disks``, so that it has a low impact on the total time
1158
  of instance creations
1159
- Added documentation for the OS parameters
1160
- Changed ``gnt-instance deactivate-disks`` so that it can work if the
1161
  hypervisor is not responding
1162
- Added display of blacklisted and hidden OS information in
1163
  ``gnt-cluster info``
1164
- Extended ``gnt-cluster verify`` to also validate hypervisor, backend,
1165
  NIC and node parameters, which might create problems with currently
1166
  invalid (but undetected) configuration files, but prevents validation
1167
  failures when unrelated parameters are modified
1168
- Changed cluster initialisation to wait for the master daemon to become
1169
  available
1170
- Expanded the RAPI interface:
1171

    
1172
  - Added config redistribution resource
1173
  - Added activation/deactivation of instance disks
1174
  - Added export of console information
1175

    
1176
- Implemented log file reopening on SIGHUP, which allows using
1177
  logrotate(8) for the Ganeti log files
1178
- Added a basic OOB helper script as an example
1179

    
1180

    
1181
Version 2.4.0 beta1
1182
-------------------
1183

    
1184
*(Released Fri, 14 Jan 2011)*
1185

    
1186
User-visible
1187
~~~~~~~~~~~~
1188

    
1189
- Fixed timezone issues when formatting timestamps
1190
- Added support for node groups, available via ``gnt-group`` and other
1191
  commands
1192
- Added out-of-band framework and management, see :doc:`design
1193
  document <design-oob>`
1194
- Removed support for roman numbers from ``gnt-node list`` and
1195
  ``gnt-instance list``.
1196
- Allowed modification of master network interface via ``gnt-cluster
1197
  modify --master-netdev``
1198
- Accept offline secondaries while shutting down instance disks
1199
- Added ``blockdev_prefix`` parameter to Xen PVM and HVM hypervisors
1200
- Added support for multiple LVM volume groups
1201
- Avoid sorting nodes for ``gnt-node list`` if specific nodes are
1202
  requested
1203
- Added commands to list available fields:
1204

    
1205
  - ``gnt-node list-fields``
1206
  - ``gnt-group list-fields``
1207
  - ``gnt-instance list-fields``
1208

    
1209
- Updated documentation and man pages
1210

    
1211
Integration
1212
~~~~~~~~~~~
1213

    
1214
- Moved ``rapi_users`` file into separate directory, now named
1215
  ``.../ganeti/rapi/users``, ``cfgupgrade`` moves the file and creates a
1216
  symlink
1217
- Added new tool for running commands on many machines,
1218
  ``tools/ganeti-listrunner``
1219
- Implemented more verbose result in ``OpInstanceConsole`` opcode, also
1220
  improving the ``gnt-instance console`` output
1221
- Allowed customisation of disk index separator at ``configure`` time
1222
- Export node group allocation policy to :doc:`iallocator <iallocator>`
1223
- Added support for non-partitioned md disks in ``lvmstrap``
1224
- Added script to gracefully power off KVM instances
1225
- Split ``utils`` module into smaller parts
1226
- Changed query operations to return more detailed information, e.g.
1227
  whether an information is unavailable due to an offline node. To use
1228
  this new functionality, the LUXI call ``Query`` must be used. Field
1229
  information is now stored by the master daemon and can be retrieved
1230
  using ``QueryFields``. Instances, nodes and groups can also be queried
1231
  using the new opcodes ``OpQuery`` and ``OpQueryFields`` (not yet
1232
  exposed via RAPI). The following commands make use of this
1233
  infrastructure change:
1234

    
1235
  - ``gnt-group list``
1236
  - ``gnt-group list-fields``
1237
  - ``gnt-node list``
1238
  - ``gnt-node list-fields``
1239
  - ``gnt-instance list``
1240
  - ``gnt-instance list-fields``
1241
  - ``gnt-debug locks``
1242

    
1243
Remote API
1244
~~~~~~~~~~
1245

    
1246
- New RAPI resources (see :doc:`rapi`):
1247

    
1248
  - ``/2/modify``
1249
  - ``/2/groups``
1250
  - ``/2/groups/[group_name]``
1251
  - ``/2/groups/[group_name]/assign-nodes``
1252
  - ``/2/groups/[group_name]/modify``
1253
  - ``/2/groups/[group_name]/rename``
1254
  - ``/2/instances/[instance_name]/disk/[disk_index]/grow``
1255

    
1256
- RAPI changes:
1257

    
1258
  - Implemented ``no_install`` for instance creation
1259
  - Implemented OS parameters for instance reinstallation, allowing
1260
    use of special settings on reinstallation (e.g. for preserving data)
1261

    
1262
Misc
1263
~~~~
1264

    
1265
- Added IPv6 support in import/export
1266
- Pause DRBD synchronization while wiping disks on instance creation
1267
- Updated unittests and QA scripts
1268
- Improved network parameters passed to KVM
1269
- Converted man pages from docbook to reStructuredText
1270

    
1271

    
1272
Version 2.3.1
1273
-------------
1274

    
1275
*(Released Mon, 20 Dec 2010)*
1276

    
1277
Released version 2.3.1~rc1 without any changes.
1278

    
1279

    
1280
Version 2.3.1 rc1
1281
-----------------
1282

    
1283
*(Released Wed, 1 Dec 2010)*
1284

    
1285
- impexpd: Disable OpenSSL compression in socat if possible (backport
1286
  from master, commit e90739d625b, see :doc:`installation guide
1287
  <install-quick>` for details)
1288
- Changed unittest coverage report to exclude test scripts
1289
- Added script to check version format
1290

    
1291

    
1292
Version 2.3.0
1293
-------------
1294

    
1295
*(Released Wed, 1 Dec 2010)*
1296

    
1297
Released version 2.3.0~rc1 without any changes.
1298

    
1299

    
1300
Version 2.3.0 rc1
1301
-----------------
1302

    
1303
*(Released Fri, 19 Nov 2010)*
1304

    
1305
A number of bugfixes and documentation updates:
1306

    
1307
- Update ganeti-os-interface documentation
1308
- Fixed a bug related to duplicate MACs or similar items which should be
1309
  unique
1310
- Fix breakage in OS state modify
1311
- Reinstall instance: disallow offline secondaries (fixes bug related to
1312
  OS changing but reinstall failing)
1313
- plus all the other fixes between 2.2.1 and 2.2.2
1314

    
1315

    
1316
Version 2.3.0 rc0
1317
-----------------
1318

    
1319
*(Released Tue, 2 Nov 2010)*
1320

    
1321
- Fixed clearing of the default iallocator using ``gnt-cluster modify``
1322
- Fixed master failover race with watcher
1323
- Fixed a bug in ``gnt-node modify`` which could lead to an inconsistent
1324
  configuration
1325
- Accept previously stopped instance for export with instance removal
1326
- Simplify and extend the environment variables for instance OS scripts
1327
- Added new node flags, ``master_capable`` and ``vm_capable``
1328
- Added optional instance disk wiping prior during allocation. This is a
1329
  cluster-wide option and can be set/modified using
1330
  ``gnt-cluster {init,modify} --prealloc-wipe-disks``.
1331
- Added IPv6 support, see :doc:`design document <design-2.3>` and
1332
  :doc:`install-quick`
1333
- Added a new watcher option (``--ignore-pause``)
1334
- Added option to ignore offline node on instance start/stop
1335
  (``--ignore-offline``)
1336
- Allow overriding OS parameters with ``gnt-instance reinstall``
1337
- Added ability to change node's secondary IP address using ``gnt-node
1338
  modify``
1339
- Implemented privilege separation for all daemons except
1340
  ``ganeti-noded``, see ``configure`` options
1341
- Complain if an instance's disk is marked faulty in ``gnt-cluster
1342
  verify``
1343
- Implemented job priorities (see ``ganeti(7)`` manpage)
1344
- Ignore failures while shutting down instances during failover from
1345
  offline node
1346
- Exit daemon's bootstrap process only once daemon is ready
1347
- Export more information via ``LUInstanceQuery``/remote API
1348
- Improved documentation, QA and unittests
1349
- RAPI daemon now watches ``rapi_users`` all the time and doesn't need a
1350
  restart if the file was created or changed
1351
- Added LUXI protocol version sent with each request and response,
1352
  allowing detection of server/client mismatches
1353
- Moved the Python scripts among gnt-* and ganeti-* into modules
1354
- Moved all code related to setting up SSH to an external script,
1355
  ``setup-ssh``
1356
- Infrastructure changes for node group support in future versions
1357

    
1358

    
1359
Version 2.2.2
1360
-------------
1361

    
1362
*(Released Fri, 19 Nov 2010)*
1363

    
1364
A few small bugs fixed, and some improvements to the build system:
1365

    
1366
- Fix documentation regarding conversion to drbd
1367
- Fix validation of parameters in cluster modify (``gnt-cluster modify
1368
  -B``)
1369
- Fix error handling in node modify with multiple changes
1370
- Allow remote imports without checked names
1371

    
1372

    
1373
Version 2.2.1
1374
-------------
1375

    
1376
*(Released Tue, 19 Oct 2010)*
1377

    
1378
- Disable SSL session ID cache in RPC client
1379

    
1380

    
1381
Version 2.2.1 rc1
1382
-----------------
1383

    
1384
*(Released Thu, 14 Oct 2010)*
1385

    
1386
- Fix interaction between Curl/GnuTLS and the Python's HTTP server
1387
  (thanks Apollon Oikonomopoulos!), finally allowing the use of Curl
1388
  with GnuTLS
1389
- Fix problems with interaction between Curl and Python's HTTP server,
1390
  resulting in increased speed in many RPC calls
1391
- Improve our release script to prevent breakage with older aclocal and
1392
  Python 2.6
1393

    
1394

    
1395
Version 2.2.1 rc0
1396
-----------------
1397

    
1398
*(Released Thu, 7 Oct 2010)*
1399

    
1400
- Fixed issue 125, replace hardcoded "xenvg" in ``gnt-cluster`` with
1401
  value retrieved from master
1402
- Added support for blacklisted or hidden OS definitions
1403
- Added simple lock monitor (accessible via (``gnt-debug locks``)
1404
- Added support for -mem-path in KVM hypervisor abstraction layer
1405
- Allow overriding instance parameters in tool for inter-cluster
1406
  instance moves (``tools/move-instance``)
1407
- Improved opcode summaries (e.g. in ``gnt-job list``)
1408
- Improve consistency of OS listing by sorting it
1409
- Documentation updates
1410

    
1411

    
1412
Version 2.2.0.1
1413
---------------
1414

    
1415
*(Released Fri, 8 Oct 2010)*
1416

    
1417
- Rebuild with a newer autotools version, to fix python 2.6 compatibility
1418

    
1419

    
1420
Version 2.2.0
1421
-------------
1422

    
1423
*(Released Mon, 4 Oct 2010)*
1424

    
1425
- Fixed regression in ``gnt-instance rename``
1426

    
1427

    
1428
Version 2.2.0 rc2
1429
-----------------
1430

    
1431
*(Released Wed, 22 Sep 2010)*
1432

    
1433
- Fixed OS_VARIANT variable for OS scripts
1434
- Fixed cluster tag operations via RAPI
1435
- Made ``setup-ssh`` exit with non-zero code if an error occurred
1436
- Disabled RAPI CA checks in watcher
1437

    
1438

    
1439
Version 2.2.0 rc1
1440
-----------------
1441

    
1442
*(Released Mon, 23 Aug 2010)*
1443

    
1444
- Support DRBD versions of the format "a.b.c.d"
1445
- Updated manpages
1446
- Re-introduce support for usage from multiple threads in RAPI client
1447
- Instance renames and modify via RAPI
1448
- Work around race condition between processing and archival in job
1449
  queue
1450
- Mark opcodes following failed one as failed, too
1451
- Job field ``lock_status`` was removed due to difficulties making it
1452
  work with the changed job queue in Ganeti 2.2; a better way to monitor
1453
  locks is expected for a later 2.2.x release
1454
- Fixed dry-run behaviour with many commands
1455
- Support ``ssh-agent`` again when adding nodes
1456
- Many additional bugfixes
1457

    
1458

    
1459
Version 2.2.0 rc0
1460
-----------------
1461

    
1462
*(Released Fri, 30 Jul 2010)*
1463

    
1464
Important change: the internal RPC mechanism between Ganeti nodes has
1465
changed from using a home-grown http library (based on the Python base
1466
libraries) to use the PycURL library. This requires that PycURL is
1467
installed on nodes. Please note that on Debian/Ubuntu, PycURL is linked
1468
against GnuTLS by default. cURL's support for GnuTLS had known issues
1469
before cURL 7.21.0 and we recommend using the latest cURL release or
1470
linking against OpenSSL. Most other distributions already link PycURL
1471
and cURL against OpenSSL. The command::
1472

    
1473
  python -c 'import pycurl; print pycurl.version'
1474

    
1475
can be used to determine the libraries PycURL and cURL are linked
1476
against.
1477

    
1478
Other significant changes:
1479

    
1480
- Rewrote much of the internals of the job queue, in order to achieve
1481
  better parallelism; this decouples job query operations from the job
1482
  processing, and it should allow much nicer behaviour of the master
1483
  daemon under load, and it also has uncovered some long-standing bugs
1484
  related to the job serialisation (now fixed)
1485
- Added a default iallocator setting to the cluster parameters,
1486
  eliminating the need to always pass nodes or an iallocator for
1487
  operations that require selection of new node(s)
1488
- Added experimental support for the LXC virtualization method
1489
- Added support for OS parameters, which allows the installation of
1490
  instances to pass parameter to OS scripts in order to customise the
1491
  instance
1492
- Added a hypervisor parameter controlling the migration type (live or
1493
  non-live), since hypervisors have various levels of reliability; this
1494
  has renamed the 'live' parameter to 'mode'
1495
- Added a cluster parameter ``reserved_lvs`` that denotes reserved
1496
  logical volumes, meaning that cluster verify will ignore them and not
1497
  flag their presence as errors
1498
- The watcher will now reset the error count for failed instances after
1499
  8 hours, thus allowing self-healing if the problem that caused the
1500
  instances to be down/fail to start has cleared in the meantime
1501
- Added a cluster parameter ``drbd_usermode_helper`` that makes Ganeti
1502
  check for, and warn, if the drbd module parameter ``usermode_helper``
1503
  is not consistent with the cluster-wide setting; this is needed to
1504
  make diagnose easier of failed drbd creations
1505
- Started adding base IPv6 support, but this is not yet
1506
  enabled/available for use
1507
- Rename operations (cluster, instance) will now return the new name,
1508
  which is especially useful if a short name was passed in
1509
- Added support for instance migration in RAPI
1510
- Added a tool to pre-configure nodes for the SSH setup, before joining
1511
  them to the cluster; this will allow in the future a simplified model
1512
  for node joining (but not yet fully enabled in 2.2); this needs the
1513
  paramiko python library
1514
- Fixed handling of name-resolving errors
1515
- Fixed consistency of job results on the error path
1516
- Fixed master-failover race condition when executed multiple times in
1517
  sequence
1518
- Fixed many bugs related to the job queue (mostly introduced during the
1519
  2.2 development cycle, so not all are impacting 2.1)
1520
- Fixed instance migration with missing disk symlinks
1521
- Fixed handling of unknown jobs in ``gnt-job archive``
1522
- And many other small fixes/improvements
1523

    
1524
Internal changes:
1525

    
1526
- Enhanced both the unittest and the QA coverage
1527
- Switched the opcode validation to a generic model, and extended the
1528
  validation to all opcode parameters
1529
- Changed more parts of the code that write shell scripts to use the
1530
  same class for this
1531
- Switched the master daemon to use the asyncore library for the Luxi
1532
  server endpoint
1533

    
1534

    
1535
Version 2.2.0 beta0
1536
-------------------
1537

    
1538
*(Released Thu, 17 Jun 2010)*
1539

    
1540
- Added tool (``move-instance``) and infrastructure to move instances
1541
  between separate clusters (see :doc:`separate documentation
1542
  <move-instance>` and :doc:`design document <design-2.2>`)
1543
- Added per-request RPC timeout
1544
- RAPI now requires a Content-Type header for requests with a body (e.g.
1545
  ``PUT`` or ``POST``) which must be set to ``application/json`` (see
1546
  :rfc:`2616` (HTTP/1.1), section 7.2.1)
1547
- ``ganeti-watcher`` attempts to restart ``ganeti-rapi`` if RAPI is not
1548
  reachable
1549
- Implemented initial support for running Ganeti daemons as separate
1550
  users, see configure-time flags ``--with-user-prefix`` and
1551
  ``--with-group-prefix`` (only ``ganeti-rapi`` is supported at this
1552
  time)
1553
- Instances can be removed after export (``gnt-backup export
1554
  --remove-instance``)
1555
- Self-signed certificates generated by Ganeti now use a 2048 bit RSA
1556
  key (instead of 1024 bit)
1557
- Added new cluster configuration file for cluster domain secret
1558
- Import/export now use SSL instead of SSH
1559
- Added support for showing estimated time when exporting an instance,
1560
  see the ``ganeti-os-interface(7)`` manpage and look for
1561
  ``EXP_SIZE_FD``
1562

    
1563

    
1564
Version 2.1.8
1565
-------------
1566

    
1567
*(Released Tue, 16 Nov 2010)*
1568

    
1569
Some more bugfixes. Unless critical bugs occur, this will be the last
1570
2.1 release:
1571

    
1572
- Fix case of MAC special-values
1573
- Fix mac checker regex
1574
- backend: Fix typo causing "out of range" error
1575
- Add missing --units in gnt-instance list man page
1576

    
1577

    
1578
Version 2.1.7
1579
-------------
1580

    
1581
*(Released Tue, 24 Aug 2010)*
1582

    
1583
Bugfixes only:
1584
  - Don't ignore secondary node silently on non-mirrored disk templates
1585
    (issue 113)
1586
  - Fix --master-netdev arg name in gnt-cluster(8) (issue 114)
1587
  - Fix usb_mouse parameter breaking with vnc_console (issue 109)
1588
  - Properly document the usb_mouse parameter
1589
  - Fix path in ganeti-rapi(8) (issue 116)
1590
  - Adjust error message when the ganeti user's .ssh directory is
1591
    missing
1592
  - Add same-node-check when changing the disk template to drbd
1593

    
1594

    
1595
Version 2.1.6
1596
-------------
1597

    
1598
*(Released Fri, 16 Jul 2010)*
1599

    
1600
Bugfixes only:
1601
  - Add an option to only select some reboot types during qa/burnin.
1602
    (on some hypervisors consequent reboots are not supported)
1603
  - Fix infrequent race condition in master failover. Sometimes the old
1604
    master ip address would be still detected as up for a short time
1605
    after it was removed, causing failover to fail.
1606
  - Decrease mlockall warnings when the ctypes module is missing. On
1607
    Python 2.4 we support running even if no ctypes module is installed,
1608
    but we were too verbose about this issue.
1609
  - Fix building on old distributions, on which man doesn't have a
1610
    --warnings option.
1611
  - Fix RAPI not to ignore the MAC address on instance creation
1612
  - Implement the old instance creation format in the RAPI client.
1613

    
1614

    
1615
Version 2.1.5
1616
-------------
1617

    
1618
*(Released Thu, 01 Jul 2010)*
1619

    
1620
A small bugfix release:
1621
  - Fix disk adoption: broken by strict --disk option checking in 2.1.4
1622
  - Fix batch-create: broken in the whole 2.1 series due to a lookup on
1623
    a non-existing option
1624
  - Fix instance create: the --force-variant option was ignored
1625
  - Improve pylint 0.21 compatibility and warnings with Python 2.6
1626
  - Fix modify node storage with non-FQDN arguments
1627
  - Fix RAPI client to authenticate under Python 2.6 when used
1628
    for more than 5 requests needing authentication
1629
  - Fix gnt-instance modify -t (storage) giving a wrong error message
1630
    when converting a non-shutdown drbd instance to plain
1631

    
1632

    
1633
Version 2.1.4
1634
-------------
1635

    
1636
*(Released Fri, 18 Jun 2010)*
1637

    
1638
A small bugfix release:
1639

    
1640
  - Fix live migration of KVM instances started with older Ganeti
1641
    versions which had fewer hypervisor parameters
1642
  - Fix gnt-instance grow-disk on down instances
1643
  - Fix an error-reporting bug during instance migration
1644
  - Better checking of the ``--net`` and ``--disk`` values, to avoid
1645
    silently ignoring broken ones
1646
  - Fix an RPC error reporting bug affecting, for example, RAPI client
1647
    users
1648
  - Fix bug triggered by different API version os-es on different nodes
1649
  - Fix a bug in instance startup with custom hvparams: OS level
1650
    parameters would fail to be applied.
1651
  - Fix the RAPI client under Python 2.6 (but more work is needed to
1652
    make it work completely well with OpenSSL)
1653
  - Fix handling of errors when resolving names from DNS
1654

    
1655

    
1656
Version 2.1.3
1657
-------------
1658

    
1659
*(Released Thu, 3 Jun 2010)*
1660

    
1661
A medium sized development cycle. Some new features, and some
1662
fixes/small improvements/cleanups.
1663

    
1664
Significant features
1665
~~~~~~~~~~~~~~~~~~~~
1666

    
1667
The node deamon now tries to mlock itself into memory, unless the
1668
``--no-mlock`` flag is passed. It also doesn't fail if it can't write
1669
its logs, and falls back to console logging. This allows emergency
1670
features such as ``gnt-node powercycle`` to work even in the event of a
1671
broken node disk (tested offlining the disk hosting the node's
1672
filesystem and dropping its memory caches; don't try this at home)
1673

    
1674
KVM: add vhost-net acceleration support. It can be tested with a new
1675
enough version of the kernel and of qemu-kvm.
1676

    
1677
KVM: Add instance chrooting feature. If you use privilege dropping for
1678
your VMs you can also now force them to chroot to an empty directory,
1679
before starting the emulated guest.
1680

    
1681
KVM: Add maximum migration bandwith and maximum downtime tweaking
1682
support (requires a new-enough version of qemu-kvm).
1683

    
1684
Cluster verify will now warn if the master node doesn't have the master
1685
ip configured on it.
1686

    
1687
Add a new (incompatible) instance creation request format to RAPI which
1688
supports all parameters (previously only a subset was supported, and it
1689
wasn't possible to extend the old format to accomodate all the new
1690
features. The old format is still supported, and a client can check for
1691
this feature, before using it, by checking for its presence in the
1692
``features`` RAPI resource.
1693

    
1694
Now with ancient latin support. Try it passing the ``--roman`` option to
1695
``gnt-instance info``, ``gnt-cluster info`` or ``gnt-node list``
1696
(requires the python-roman module to be installed, in order to work).
1697

    
1698
Other changes
1699
~~~~~~~~~~~~~
1700

    
1701
As usual many internal code refactorings, documentation updates, and
1702
such. Among others:
1703

    
1704
  - Lots of improvements and cleanups to the experimental Remote API
1705
    (RAPI) client library.
1706
  - A new unit test suite for the core daemon libraries.
1707
  - A fix to creating missing directories makes sure the umask is not
1708
    applied anymore. This enforces the same directory permissions
1709
    everywhere.
1710
  - Better handling terminating daemons with ctrl+c (used when running
1711
    them in debugging mode).
1712
  - Fix a race condition in live migrating a KVM instance, when stat()
1713
    on the old proc status file returned EINVAL, which is an unexpected
1714
    value.
1715
  - Fixed manpage checking with newer man and utf-8 charachters. But now
1716
    you need the en_US.UTF-8 locale enabled to build Ganeti from git.
1717

    
1718

    
1719
Version 2.1.2.1
1720
---------------
1721

    
1722
*(Released Fri, 7 May 2010)*
1723

    
1724
Fix a bug which prevented untagged KVM instances from starting.
1725

    
1726

    
1727
Version 2.1.2
1728
-------------
1729

    
1730
*(Released Fri, 7 May 2010)*
1731

    
1732
Another release with a long development cycle, during which many
1733
different features were added.
1734

    
1735
Significant features
1736
~~~~~~~~~~~~~~~~~~~~
1737

    
1738
The KVM hypervisor now can run the individual instances as non-root, to
1739
reduce the impact of a VM being hijacked due to bugs in the
1740
hypervisor. It is possible to run all instances as a single (non-root)
1741
user, to manually specify a user for each instance, or to dynamically
1742
allocate a user out of a cluster-wide pool to each instance, with the
1743
guarantee that no two instances will run under the same user ID on any
1744
given node.
1745

    
1746
An experimental RAPI client library, that can be used standalone
1747
(without the other Ganeti libraries), is provided in the source tree as
1748
``lib/rapi/client.py``. Note this client might change its interface in
1749
the future, as we iterate on its capabilities.
1750

    
1751
A new command, ``gnt-cluster renew-crypto`` has been added to easily
1752
replace the cluster's certificates and crypto keys. This might help in
1753
case they have been compromised, or have simply expired.
1754

    
1755
A new disk option for instance creation has been added that allows one
1756
to "adopt" currently existing logical volumes, with data
1757
preservation. This should allow easier migration to Ganeti from
1758
unmanaged (or managed via other software) instances.
1759

    
1760
Another disk improvement is the possibility to convert between redundant
1761
(DRBD) and plain (LVM) disk configuration for an instance. This should
1762
allow better scalability (starting with one node and growing the
1763
cluster, or shrinking a two-node cluster to one node).
1764

    
1765
A new feature that could help with automated node failovers has been
1766
implemented: if a node sees itself as offline (by querying the master
1767
candidates), it will try to shutdown (hard) all instances and any active
1768
DRBD devices. This reduces the risk of duplicate instances if an
1769
external script automatically failovers the instances on such nodes. To
1770
enable this, the cluster parameter ``maintain_node_health`` should be
1771
enabled; in the future this option (per the name) will enable other
1772
automatic maintenance features.
1773

    
1774
Instance export/import now will reuse the original instance
1775
specifications for all parameters; that means exporting an instance,
1776
deleting it and the importing it back should give an almost identical
1777
instance. Note that the default import behaviour has changed from
1778
before, where it created only one NIC; now it recreates the original
1779
number of NICs.
1780

    
1781
Cluster verify has added a few new checks: SSL certificates validity,
1782
/etc/hosts consistency across the cluster, etc.
1783

    
1784
Other changes
1785
~~~~~~~~~~~~~
1786

    
1787
As usual, many internal changes were done, documentation fixes,
1788
etc. Among others:
1789

    
1790
- Fixed cluster initialization with disabled cluster storage (regression
1791
  introduced in 2.1.1)
1792
- File-based storage supports growing the disks
1793
- Fixed behaviour of node role changes
1794
- Fixed cluster verify for some corner cases, plus a general rewrite of
1795
  cluster verify to allow future extension with more checks
1796
- Fixed log spamming by watcher and node daemon (regression introduced
1797
  in 2.1.1)
1798
- Fixed possible validation issues when changing the list of enabled
1799
  hypervisors
1800
- Fixed cleanup of /etc/hosts during node removal
1801
- Fixed RAPI response for invalid methods
1802
- Fixed bug with hashed passwords in ``ganeti-rapi`` daemon
1803
- Multiple small improvements to the KVM hypervisor (VNC usage, booting
1804
  from ide disks, etc.)
1805
- Allow OS changes without re-installation (to record a changed OS
1806
  outside of Ganeti, or to allow OS renames)
1807
- Allow instance creation without OS installation (useful for example if
1808
  the OS will be installed manually, or restored from a backup not in
1809
  Ganeti format)
1810
- Implemented option to make cluster ``copyfile`` use the replication
1811
  network
1812
- Added list of enabled hypervisors to ssconf (possibly useful for
1813
  external scripts)
1814
- Added a new tool (``tools/cfgupgrade12``) that allows upgrading from
1815
  1.2 clusters
1816
- A partial form of node re-IP is possible via node readd, which now
1817
  allows changed node primary IP
1818
- Command line utilities now show an informational message if the job is
1819
  waiting for a lock
1820
- The logs of the master daemon now show the PID/UID/GID of the
1821
  connected client
1822

    
1823

    
1824
Version 2.1.1
1825
-------------
1826

    
1827
*(Released Fri, 12 Mar 2010)*
1828

    
1829
During the 2.1.0 long release candidate cycle, a lot of improvements and
1830
changes have accumulated with were released later as 2.1.1.
1831

    
1832
Major changes
1833
~~~~~~~~~~~~~
1834

    
1835
The node evacuate command (``gnt-node evacuate``) was significantly
1836
rewritten, and as such the IAllocator protocol was changed - a new
1837
request type has been added. This unfortunate change during a stable
1838
series is designed to improve performance of node evacuations; on
1839
clusters with more than about five nodes and which are well-balanced,
1840
evacuation should proceed in parallel for all instances of the node
1841
being evacuated. As such, any existing IAllocator scripts need to be
1842
updated, otherwise the above command will fail due to the unknown
1843
request. The provided "dumb" allocator has not been updated; but the
1844
ganeti-htools package supports the new protocol since version 0.2.4.
1845

    
1846
Another important change is increased validation of node and instance
1847
names. This might create problems in special cases, if invalid host
1848
names are being used.
1849

    
1850
Also, a new layer of hypervisor parameters has been added, that sits at
1851
OS level between the cluster defaults and the instance ones. This allows
1852
customisation of virtualization parameters depending on the installed
1853
OS. For example instances with OS 'X' may have a different KVM kernel
1854
(or any other parameter) than the cluster defaults. This is intended to
1855
help managing a multiple OSes on the same cluster, without manual
1856
modification of each instance's parameters.
1857

    
1858
A tool for merging clusters, ``cluster-merge``, has been added in the
1859
tools sub-directory.
1860

    
1861
Bug fixes
1862
~~~~~~~~~
1863

    
1864
- Improved the int/float conversions that should make the code more
1865
  robust in face of errors from the node daemons
1866
- Fixed the remove node code in case of internal configuration errors
1867
- Fixed the node daemon behaviour in face of inconsistent queue
1868
  directory (e.g. read-only file-system where we can't open the files
1869
  read-write, etc.)
1870
- Fixed the behaviour of gnt-node modify for master candidate demotion;
1871
  now it either aborts cleanly or, if given the new "auto_promote"
1872
  parameter, will automatically promote other nodes as needed
1873
- Fixed compatibility with (unreleased yet) Python 2.6.5 that would
1874
  completely prevent Ganeti from working
1875
- Fixed bug for instance export when not all disks were successfully
1876
  exported
1877
- Fixed behaviour of node add when the new node is slow in starting up
1878
  the node daemon
1879
- Fixed handling of signals in the LUXI client, which should improve
1880
  behaviour of command-line scripts
1881
- Added checks for invalid node/instance names in the configuration (now
1882
  flagged during cluster verify)
1883
- Fixed watcher behaviour for disk activation errors
1884
- Fixed two potentially endless loops in http library, which led to the
1885
  RAPI daemon hanging and consuming 100% CPU in some cases
1886
- Fixed bug in RAPI daemon related to hashed passwords
1887
- Fixed bug for unintended qemu-level bridging of multi-NIC KVM
1888
  instances
1889
- Enhanced compatibility with non-Debian OSes, but not using absolute
1890
  path in some commands and allowing customisation of the ssh
1891
  configuration directory
1892
- Fixed possible future issue with new Python versions by abiding to the
1893
  proper use of ``__slots__`` attribute on classes
1894
- Added checks that should prevent directory traversal attacks
1895
- Many documentation fixes based on feedback from users
1896

    
1897
New features
1898
~~~~~~~~~~~~
1899

    
1900
- Added an "early_release" more for instance replace disks and node
1901
  evacuate, where we release locks earlier and thus allow higher
1902
  parallelism within the cluster
1903
- Added watcher hooks, intended to allow the watcher to restart other
1904
  daemons (e.g. from the ganeti-nbma project), but they can be used of
1905
  course for any other purpose
1906
- Added a compile-time disable for DRBD barriers, to increase
1907
  performance if the administrator trusts the power supply or the
1908
  storage system to not lose writes
1909
- Added the option of using syslog for logging instead of, or in
1910
  addition to, Ganeti's own log files
1911
- Removed boot restriction for paravirtual NICs for KVM, recent versions
1912
  can indeed boot from a paravirtual NIC
1913
- Added a generic debug level for many operations; while this is not
1914
  used widely yet, it allows one to pass the debug value all the way to
1915
  the OS scripts
1916
- Enhanced the hooks environment for instance moves (failovers,
1917
  migrations) where the primary/secondary nodes changed during the
1918
  operation, by adding {NEW,OLD}_{PRIMARY,SECONDARY} vars
1919
- Enhanced data validations for many user-supplied values; one important
1920
  item is the restrictions imposed on instance and node names, which
1921
  might reject some (invalid) host names
1922
- Add a configure-time option to disable file-based storage, if it's not
1923
  needed; this allows greater security separation between the master
1924
  node and the other nodes from the point of view of the inter-node RPC
1925
  protocol
1926
- Added user notification in interactive tools if job is waiting in the
1927
  job queue or trying to acquire locks
1928
- Added log messages when a job is waiting for locks
1929
- Added filtering by node tags in instance operations which admit
1930
  multiple instances (start, stop, reboot, reinstall)
1931
- Added a new tool for cluster mergers, ``cluster-merge``
1932
- Parameters from command line which are of the form ``a=b,c=d`` can now
1933
  use backslash escapes to pass in values which contain commas,
1934
  e.g. ``a=b\\c,d=e`` where the 'a' parameter would get the value
1935
  ``b,c``
1936
- For KVM, the instance name is the first parameter passed to KVM, so
1937
  that it's more visible in the process list
1938

    
1939

    
1940
Version 2.1.0
1941
-------------
1942

    
1943
*(Released Tue, 2 Mar 2010)*
1944

    
1945
Ganeti 2.1 brings many improvements with it. Major changes:
1946

    
1947
- Added infrastructure to ease automated disk repairs
1948
- Added new daemon to export configuration data in a cheaper way than
1949
  using the remote API
1950
- Instance NICs can now be routed instead of being associated with a
1951
  networking bridge
1952
- Improved job locking logic to reduce impact of jobs acquiring multiple
1953
  locks waiting for other long-running jobs
1954

    
1955
In-depth implementation details can be found in the Ganeti 2.1 design
1956
document.
1957

    
1958
Details
1959
~~~~~~~
1960

    
1961
- Added chroot hypervisor
1962
- Added more options to xen-hvm hypervisor (``kernel_path`` and
1963
  ``device_model``)
1964
- Added more options to xen-pvm hypervisor (``use_bootloader``,
1965
  ``bootloader_path`` and ``bootloader_args``)
1966
- Added the ``use_localtime`` option for the xen-hvm and kvm
1967
  hypervisors, and the default value for this has changed to false (in
1968
  2.0 xen-hvm always enabled it)
1969
- Added luxi call to submit multiple jobs in one go
1970
- Added cluster initialization option to not modify ``/etc/hosts``
1971
  file on nodes
1972
- Added network interface parameters
1973
- Added dry run mode to some LUs
1974
- Added RAPI resources:
1975

    
1976
  - ``/2/instances/[instance_name]/info``
1977
  - ``/2/instances/[instance_name]/replace-disks``
1978
  - ``/2/nodes/[node_name]/evacuate``
1979
  - ``/2/nodes/[node_name]/migrate``
1980
  - ``/2/nodes/[node_name]/role``
1981
  - ``/2/nodes/[node_name]/storage``
1982
  - ``/2/nodes/[node_name]/storage/modify``
1983
  - ``/2/nodes/[node_name]/storage/repair``
1984

    
1985
- Added OpCodes to evacuate or migrate all instances on a node
1986
- Added new command to list storage elements on nodes (``gnt-node
1987
  list-storage``) and modify them (``gnt-node modify-storage``)
1988
- Added new ssconf files with master candidate IP address
1989
  (``ssconf_master_candidates_ips``), node primary IP address
1990
  (``ssconf_node_primary_ips``) and node secondary IP address
1991
  (``ssconf_node_secondary_ips``)
1992
- Added ``ganeti-confd`` and a client library to query the Ganeti
1993
  configuration via UDP
1994
- Added ability to run hooks after cluster initialization and before
1995
  cluster destruction
1996
- Added automatic mode for disk replace (``gnt-instance replace-disks
1997
  --auto``)
1998
- Added ``gnt-instance recreate-disks`` to re-create (empty) disks
1999
  after catastrophic data-loss
2000
- Added ``gnt-node repair-storage`` command to repair damaged LVM volume
2001
  groups
2002
- Added ``gnt-instance move`` command to move instances
2003
- Added ``gnt-cluster watcher`` command to control watcher
2004
- Added ``gnt-node powercycle`` command to powercycle nodes
2005
- Added new job status field ``lock_status``
2006
- Added parseable error codes to cluster verification (``gnt-cluster
2007
  verify --error-codes``) and made output less verbose (use
2008
  ``--verbose`` to restore previous behaviour)
2009
- Added UUIDs to the main config entities (cluster, nodes, instances)
2010
- Added support for OS variants
2011
- Added support for hashed passwords in the Ganeti remote API users file
2012
  (``rapi_users``)
2013
- Added option to specify maximum timeout on instance shutdown
2014
- Added ``--no-ssh-init`` option to ``gnt-cluster init``
2015
- Added new helper script to start and stop Ganeti daemons
2016
  (``daemon-util``), with the intent to reduce the work necessary to
2017
  adjust Ganeti for non-Debian distributions and to start/stop daemons
2018
  from one place
2019
- Added more unittests
2020
- Fixed critical bug in ganeti-masterd startup
2021
- Removed the configure-time ``kvm-migration-port`` parameter, this is
2022
  now customisable at the cluster level for both the KVM and Xen
2023
  hypervisors using the new ``migration_port`` parameter
2024
- Pass ``INSTANCE_REINSTALL`` variable to OS installation script when
2025
  reinstalling an instance
2026
- Allowed ``@`` in tag names
2027
- Migrated to Sphinx (http://sphinx.pocoo.org/) for documentation
2028
- Many documentation updates
2029
- Distribute hypervisor files on ``gnt-cluster redist-conf``
2030
- ``gnt-instance reinstall`` can now reinstall multiple instances
2031
- Updated many command line parameters
2032
- Introduced new OS API version 15
2033
- No longer support a default hypervisor
2034
- Treat virtual LVs as inexistent
2035
- Improved job locking logic to reduce lock contention
2036
- Match instance and node names case insensitively
2037
- Reimplemented bash completion script to be more complete
2038
- Improved burnin
2039

    
2040

    
2041
Version 2.0.6
2042
-------------
2043

    
2044
*(Released Thu, 4 Feb 2010)*
2045

    
2046
- Fix cleaner behaviour on nodes not in a cluster (Debian bug 568105)
2047
- Fix a string formatting bug
2048
- Improve safety of the code in some error paths
2049
- Improve data validation in the master of values returned from nodes
2050

    
2051

    
2052
Version 2.0.5
2053
-------------
2054

    
2055
*(Released Thu, 17 Dec 2009)*
2056

    
2057
- Fix security issue due to missing validation of iallocator names; this
2058
  allows local and remote execution of arbitrary executables
2059
- Fix failure of gnt-node list during instance removal
2060
- Ship the RAPI documentation in the archive
2061

    
2062

    
2063
Version 2.0.4
2064
-------------
2065

    
2066
*(Released Wed, 30 Sep 2009)*
2067

    
2068
- Fixed many wrong messages
2069
- Fixed a few bugs related to the locking library
2070
- Fixed MAC checking at instance creation time
2071
- Fixed a DRBD parsing bug related to gaps in /proc/drbd
2072
- Fixed a few issues related to signal handling in both daemons and
2073
  scripts
2074
- Fixed the example startup script provided
2075
- Fixed insserv dependencies in the example startup script (patch from
2076
  Debian)
2077
- Fixed handling of drained nodes in the iallocator framework
2078
- Fixed handling of KERNEL_PATH parameter for xen-hvm (Debian bug
2079
  #528618)
2080
- Fixed error related to invalid job IDs in job polling
2081
- Fixed job/opcode persistence on unclean master shutdown
2082
- Fixed handling of partial job processing after unclean master
2083
  shutdown
2084
- Fixed error reporting from LUs, previously all errors were converted
2085
  into execution errors
2086
- Fixed error reporting from burnin
2087
- Decreased significantly the memory usage of the job queue
2088
- Optimised slightly multi-job submission
2089
- Optimised slightly opcode loading
2090
- Backported the multi-job submit framework from the development
2091
  branch; multi-instance start and stop should be faster
2092
- Added script to clean archived jobs after 21 days; this will reduce
2093
  the size of the queue directory
2094
- Added some extra checks in disk size tracking
2095
- Added an example ethers hook script
2096
- Added a cluster parameter that prevents Ganeti from modifying of
2097
  /etc/hosts
2098
- Added more node information to RAPI responses
2099
- Added a ``gnt-job watch`` command that allows following the ouput of a
2100
  job
2101
- Added a bind-address option to ganeti-rapi
2102
- Added more checks to the configuration verify
2103
- Enhanced the burnin script such that some operations can be retried
2104
  automatically
2105
- Converted instance reinstall to multi-instance model
2106

    
2107

    
2108
Version 2.0.3
2109
-------------
2110

    
2111
*(Released Fri, 7 Aug 2009)*
2112

    
2113
- Added ``--ignore-size`` to the ``gnt-instance activate-disks`` command
2114
  to allow using the pre-2.0.2 behaviour in activation, if any existing
2115
  instances have mismatched disk sizes in the configuration
2116
- Added ``gnt-cluster repair-disk-sizes`` command to check and update
2117
  any configuration mismatches for disk sizes
2118
- Added ``gnt-master cluste-failover --no-voting`` to allow master
2119
  failover to work on two-node clusters
2120
- Fixed the ``--net`` option of ``gnt-backup import``, which was
2121
  unusable
2122
- Fixed detection of OS script errors in ``gnt-backup export``
2123
- Fixed exit code of ``gnt-backup export``
2124

    
2125

    
2126
Version 2.0.2
2127
-------------
2128

    
2129
*(Released Fri, 17 Jul 2009)*
2130

    
2131
- Added experimental support for stripped logical volumes; this should
2132
  enhance performance but comes with a higher complexity in the block
2133
  device handling; stripping is only enabled when passing
2134
  ``--with-lvm-stripecount=N`` to ``configure``, but codepaths are
2135
  affected even in the non-stripped mode
2136
- Improved resiliency against transient failures at the end of DRBD
2137
  resyncs, and in general of DRBD resync checks
2138
- Fixed a couple of issues with exports and snapshot errors
2139
- Fixed a couple of issues in instance listing
2140
- Added display of the disk size in ``gnt-instance info``
2141
- Fixed checking for valid OSes in instance creation
2142
- Fixed handling of the "vcpus" parameter in instance listing and in
2143
  general of invalid parameters
2144
- Fixed http server library, and thus RAPI, to handle invalid
2145
  username/password combinations correctly; this means that now they
2146
  report unauthorized for queries too, not only for modifications,
2147
  allowing earlier detect of configuration problems
2148
- Added a new "role" node list field, equivalent to the master/master
2149
  candidate/drained/offline flags combinations
2150
- Fixed cluster modify and changes of candidate pool size
2151
- Fixed cluster verify error messages for wrong files on regular nodes
2152
- Fixed a couple of issues with node demotion from master candidate role
2153
- Fixed node readd issues
2154
- Added non-interactive mode for ``ganeti-masterd --no-voting`` startup
2155
- Added a new ``--no-voting`` option for masterfailover to fix failover
2156
  on two-nodes clusters when the former master node is unreachable
2157
- Added instance reinstall over RAPI
2158

    
2159

    
2160
Version 2.0.1
2161
-------------
2162

    
2163
*(Released Tue, 16 Jun 2009)*
2164

    
2165
- added ``-H``/``-B`` startup parameters to ``gnt-instance``, which will
2166
  allow re-adding the start in single-user option (regression from 1.2)
2167
- the watcher writes the instance status to a file, to allow monitoring
2168
  to report the instance status (from the master) based on cached
2169
  results of the watcher's queries; while this can get stale if the
2170
  watcher is being locked due to other work on the cluster, this is
2171
  still an improvement
2172
- the watcher now also restarts the node daemon and the rapi daemon if
2173
  they died
2174
- fixed the watcher to handle full and drained queue cases
2175
- hooks export more instance data in the environment, which helps if
2176
  hook scripts need to take action based on the instance's properties
2177
  (no longer need to query back into ganeti)
2178
- instance failovers when the instance is stopped do not check for free
2179
  RAM, so that failing over a stopped instance is possible in low memory
2180
  situations
2181
- rapi uses queries for tags instead of jobs (for less job traffic), and
2182
  for cluster tags it won't talk to masterd at all but read them from
2183
  ssconf
2184
- a couple of error handling fixes in RAPI
2185
- drbd handling: improved the error handling of inconsistent disks after
2186
  resync to reduce the frequency of "there are some degraded disks for
2187
  this instance" messages
2188
- fixed a bug in live migration when DRBD doesn't want to reconnect (the
2189
  error handling path called a wrong function name)
2190

    
2191

    
2192
Version 2.0.0
2193
-------------
2194

    
2195
*(Released Wed, 27 May 2009)*
2196

    
2197
- no changes from rc5
2198

    
2199

    
2200
Version 2.0 rc5
2201
---------------
2202

    
2203
*(Released Wed, 20 May 2009)*
2204

    
2205
- fix a couple of bugs (validation, argument checks)
2206
- fix ``gnt-cluster getmaster`` on non-master nodes (regression)
2207
- some small improvements to RAPI and IAllocator
2208
- make watcher automatically start the master daemon if down
2209

    
2210

    
2211
Version 2.0 rc4
2212
---------------
2213

    
2214
*(Released Mon, 27 Apr 2009)*
2215

    
2216
- change the OS list to not require locks; this helps with big clusters
2217
- fix ``gnt-cluster verify`` and ``gnt-cluster verify-disks`` when the
2218
  volume group is broken
2219
- ``gnt-instance info``, without any arguments, doesn't run for all
2220
  instances anymore; either pass ``--all`` or pass the desired
2221
  instances; this helps against mistakes on big clusters where listing
2222
  the information for all instances takes a long time
2223
- miscellaneous doc and man pages fixes
2224

    
2225

    
2226
Version 2.0 rc3
2227
---------------
2228

    
2229
*(Released Wed, 8 Apr 2009)*
2230

    
2231
- Change the internal locking model of some ``gnt-node`` commands, in
2232
  order to reduce contention (and blocking of master daemon) when
2233
  batching many creation/reinstall jobs
2234
- Fixes to Xen soft reboot
2235
- No longer build documentation at build time, instead distribute it in
2236
  the archive, in order to reduce the need for the whole docbook/rst
2237
  toolchains
2238

    
2239

    
2240
Version 2.0 rc2
2241
---------------
2242

    
2243
*(Released Fri, 27 Mar 2009)*
2244

    
2245
- Now the cfgupgrade scripts works and can upgrade 1.2.7 clusters to 2.0
2246
- Fix watcher startup sequence, improves the behaviour of busy clusters
2247
- Some other fixes in ``gnt-cluster verify``, ``gnt-instance
2248
  replace-disks``, ``gnt-instance add``, ``gnt-cluster queue``, KVM VNC
2249
  bind address and other places
2250
- Some documentation fixes and updates
2251

    
2252

    
2253
Version 2.0 rc1
2254
---------------
2255

    
2256
*(Released Mon, 2 Mar 2009)*
2257

    
2258
- More documentation updates, now all docs should be more-or-less
2259
  up-to-date
2260
- A couple of small fixes (mixed hypervisor clusters, offline nodes,
2261
  etc.)
2262
- Added a customizable HV_KERNEL_ARGS hypervisor parameter (for Xen PVM
2263
  and KVM)
2264
- Fix an issue related to $libdir/run/ganeti and cluster creation
2265

    
2266

    
2267
Version 2.0 beta2
2268
-----------------
2269

    
2270
*(Released Thu, 19 Feb 2009)*
2271

    
2272
- Xen PVM and KVM have switched the default value for the instance root
2273
  disk to the first partition on the first drive, instead of the whole
2274
  drive; this means that the OS installation scripts must be changed
2275
  accordingly
2276
- Man pages have been updated
2277
- RAPI has been switched by default to HTTPS, and the exported functions
2278
  should all work correctly
2279
- RAPI v1 has been removed
2280
- Many improvements to the KVM hypervisor
2281
- Block device errors are now better reported
2282
- Many other bugfixes and small improvements
2283

    
2284

    
2285
Version 2.0 beta1
2286
-----------------
2287

    
2288
*(Released Mon, 26 Jan 2009)*
2289

    
2290
- Version 2 is a general rewrite of the code and therefore the
2291
  differences are too many to list, see the design document for 2.0 in
2292
  the ``doc/`` subdirectory for more details
2293
- In this beta version there is not yet a migration path from 1.2 (there
2294
  will be one in the final 2.0 release)
2295
- A few significant changes are:
2296

    
2297
  - all commands are executed by a daemon (``ganeti-masterd``) and the
2298
    various ``gnt-*`` commands are just front-ends to it
2299
  - all the commands are entered into, and executed from a job queue,
2300
    see the ``gnt-job(8)`` manpage
2301
  - the RAPI daemon supports read-write operations, secured by basic
2302
    HTTP authentication on top of HTTPS
2303
  - DRBD version 0.7 support has been removed, DRBD 8 is the only
2304
    supported version (when migrating from Ganeti 1.2 to 2.0, you need
2305
    to migrate to DRBD 8 first while still running Ganeti 1.2)
2306
  - DRBD devices are using statically allocated minor numbers, which
2307
    will be assigned to existing instances during the migration process
2308
  - there is support for both Xen PVM and Xen HVM instances running on
2309
    the same cluster
2310
  - KVM virtualization is supported too
2311
  - file-based storage has been implemented, which means that it is
2312
    possible to run the cluster without LVM and DRBD storage, for
2313
    example using a shared filesystem exported from shared storage (and
2314
    still have live migration)
2315

    
2316

    
2317
Version 1.2.7
2318
-------------
2319

    
2320
*(Released Tue, 13 Jan 2009)*
2321

    
2322
- Change the default reboot type in ``gnt-instance reboot`` to "hard"
2323
- Reuse the old instance mac address by default on instance import, if
2324
  the instance name is the same.
2325
- Handle situations in which the node info rpc returns incomplete
2326
  results (issue 46)
2327
- Add checks for tcp/udp ports collisions in ``gnt-cluster verify``
2328
- Improved version of batcher:
2329

    
2330
  - state file support
2331
  - instance mac address support
2332
  - support for HVM clusters/instances
2333

    
2334
- Add an option to show the number of cpu sockets and nodes in
2335
  ``gnt-node list``
2336
- Support OSes that handle more than one version of the OS api (but do
2337
  not change the current API in any other way)
2338
- Fix ``gnt-node migrate``
2339
- ``gnt-debug`` man page
2340
- Fixes various more typos and small issues
2341
- Increase disk resync maximum speed to 60MB/s (from 30MB/s)
2342

    
2343

    
2344
Version 1.2.6
2345
-------------
2346

    
2347
*(Released Wed, 24 Sep 2008)*
2348

    
2349
- new ``--hvm-nic-type`` and ``--hvm-disk-type`` flags to control the
2350
  type of disk exported to fully virtualized instances.
2351
- provide access to the serial console of HVM instances
2352
- instance auto_balance flag, set by default. If turned off it will
2353
  avoid warnings on cluster verify if there is not enough memory to fail
2354
  over an instance. in the future it will prevent automatically failing
2355
  it over when we will support that.
2356
- batcher tool for instance creation, see ``tools/README.batcher``
2357
- ``gnt-instance reinstall --select-os`` to interactively select a new
2358
  operating system when reinstalling an instance.
2359
- when changing the memory amount on instance modify a check has been
2360
  added that the instance will be able to start. also warnings are
2361
  emitted if the instance will not be able to fail over, if auto_balance
2362
  is true.
2363
- documentation fixes
2364
- sync fields between ``gnt-instance list/modify/add/import``
2365
- fix a race condition in drbd when the sync speed was set after giving
2366
  the device a remote peer.
2367

    
2368

    
2369
Version 1.2.5
2370
-------------
2371

    
2372
*(Released Tue, 22 Jul 2008)*
2373

    
2374
- note: the allowed size and number of tags per object were reduced
2375
- fix a bug in ``gnt-cluster verify`` with inconsistent volume groups
2376
- fixed twisted 8.x compatibility
2377
- fixed ``gnt-instance replace-disks`` with iallocator
2378
- add TCP keepalives on twisted connections to detect restarted nodes
2379
- disk increase support, see ``gnt-instance grow-disk``
2380
- implement bulk node/instance query for RAPI
2381
- add tags in node/instance listing (optional)
2382
- experimental migration (and live migration) support, read the man page
2383
  for ``gnt-instance migrate``
2384
- the ``ganeti-watcher`` logs are now timestamped, and the watcher also
2385
  has some small improvements in handling its state file
2386

    
2387

    
2388
Version 1.2.4
2389
-------------
2390

    
2391
*(Released Fri, 13 Jun 2008)*
2392

    
2393
- Experimental readonly, REST-based remote API implementation;
2394
  automatically started on master node, TCP port 5080, if enabled by
2395
  ``--enable-rapi`` parameter to configure script.
2396
- Instance allocator support. Add and import instance accept a
2397
  ``--iallocator`` parameter, and call that instance allocator to decide
2398
  which node to use for the instance. The iallocator document describes
2399
  what's expected from an allocator script.
2400
- ``gnt-cluster verify`` N+1 memory redundancy checks: Unless passed the
2401
  ``--no-nplus1-mem`` option ``gnt-cluster verify`` now checks that if a
2402
  node is lost there is still enough memory to fail over the instances
2403
  that reside on it.
2404
- ``gnt-cluster verify`` hooks: it is now possible to add post-hooks to
2405
  ``gnt-cluster verify``, to check for site-specific compliance. All the
2406
  hooks will run, and their output, if any, will be displayed. Any
2407
  failing hook will make the verification return an error value.
2408
- ``gnt-cluster verify`` now checks that its peers are reachable on the
2409
  primary and secondary interfaces
2410
- ``gnt-node add`` now supports the ``--readd`` option, to readd a node
2411
  that is still declared as part of the cluster and has failed.
2412
- ``gnt-* list`` commands now accept a new ``-o +field`` way of
2413
  specifying output fields, that just adds the chosen fields to the
2414
  default ones.
2415
- ``gnt-backup`` now has a new ``remove`` command to delete an existing
2416
  export from the filesystem.
2417
- New per-instance parameters hvm_acpi, hvm_pae and hvm_cdrom_image_path
2418
  have been added. Using them you can enable/disable acpi and pae
2419
  support, and specify a path for a cd image to be exported to the
2420
  instance. These parameters as the name suggest only work on HVM
2421
  clusters.
2422
- When upgrading an HVM cluster to Ganeti 1.2.4, the values for ACPI and
2423
  PAE support will be set to the previously hardcoded values, but the
2424
  (previously hardcoded) path to the CDROM ISO image will be unset and
2425
  if required, needs to be set manually with ``gnt-instance modify``
2426
  after the upgrade.
2427
- The address to which an instance's VNC console is bound is now
2428
  selectable per-instance, rather than being cluster wide. Of course
2429
  this only applies to instances controlled via VNC, so currently just
2430
  applies to HVM clusters.
2431

    
2432

    
2433
Version 1.2.3
2434
-------------
2435

    
2436
*(Released Mon, 18 Feb 2008)*
2437

    
2438
- more tweaks to the disk activation code (especially helpful for DRBD)
2439
- change the default ``gnt-instance list`` output format, now there is
2440
  one combined status field (see the manpage for the exact values this
2441
  field will have)
2442
- some more fixes for the mac export to hooks change
2443
- make Ganeti not break with DRBD 8.2.x (which changed the version
2444
  format in ``/proc/drbd``) (issue 24)
2445
- add an upgrade tool from "remote_raid1" disk template to "drbd" disk
2446
  template, allowing migration from DRBD0.7+MD to DRBD8
2447

    
2448

    
2449
Version 1.2.2
2450
-------------
2451

    
2452
*(Released Wed, 30 Jan 2008)*
2453

    
2454
- fix ``gnt-instance modify`` breakage introduced in 1.2.1 with the HVM
2455
  support (issue 23)
2456
- add command aliases infrastructure and a few aliases
2457
- allow listing of VCPUs in the ``gnt-instance list`` and improve the
2458
  man pages and the ``--help`` option of ``gnt-node
2459
  list``/``gnt-instance list``
2460
- fix ``gnt-backup list`` with down nodes (issue 21)
2461
- change the tools location (move from $pkgdatadir to $pkglibdir/tools)
2462
- fix the dist archive and add a check for including svn/git files in
2463
  the future
2464
- some developer-related changes: improve the burnin and the QA suite,
2465
  add an upload script for testing during development
2466

    
2467

    
2468
Version 1.2.1
2469
-------------
2470

    
2471
*(Released Wed, 16 Jan 2008)*
2472

    
2473
- experimental HVM support, read the install document, section
2474
  "Initializing the cluster"
2475
- allow for the PVM hypervisor per-instance kernel and initrd paths
2476
- add a new command ``gnt-cluster verify-disks`` which uses a new
2477
  algorithm to improve the reconnection of the DRBD pairs if the device
2478
  on the secondary node has gone away
2479
- make logical volume code auto-activate LVs at disk activation time
2480
- slightly improve the speed of activating disks
2481
- allow specification of the MAC address at instance creation time, and
2482
  changing it later via ``gnt-instance modify``
2483
- fix handling of external commands that generate lots of output on
2484
  stderr
2485
- update documentation with regard to minimum version of DRBD8 supported
2486

    
2487

    
2488
Version 1.2.0
2489
-------------
2490

    
2491
*(Released Tue, 4 Dec 2007)*
2492

    
2493
- Log the ``xm create`` output to the node daemon log on failure (to
2494
  help diagnosing the error)
2495
- In debug mode, log all external commands output if failed to the logs
2496
- Change parsing of lvm commands to ignore stderr
2497

    
2498

    
2499
Version 1.2 beta3
2500
-----------------
2501

    
2502
*(Released Wed, 28 Nov 2007)*
2503

    
2504
- Another round of updates to the DRBD 8 code to deal with more failures
2505
  in the replace secondary node operation
2506
- Some more logging of failures in disk operations (lvm, drbd)
2507
- A few documentation updates
2508
- QA updates
2509

    
2510

    
2511
Version 1.2 beta2
2512
-----------------
2513

    
2514
*(Released Tue, 13 Nov 2007)*
2515

    
2516
- Change configuration file format from Python's Pickle to JSON.
2517
  Upgrading is possible using the cfgupgrade utility.
2518
- Add support for DRBD 8.0 (new disk template ``drbd``) which allows for
2519
  faster replace disks and is more stable (DRBD 8 has many improvements
2520
  compared to DRBD 0.7)
2521
- Added command line tags support (see man pages for ``gnt-instance``,
2522
  ``gnt-node``, ``gnt-cluster``)
2523
- Added instance rename support
2524
- Added multi-instance startup/shutdown
2525
- Added cluster rename support
2526
- Added ``gnt-node evacuate`` to simplify some node operations
2527
- Added instance reboot operation that can speedup reboot as compared to
2528
  stop and start
2529
- Soften the requirement that hostnames are in FQDN format
2530
- The ``ganeti-watcher`` now activates drbd pairs after secondary node
2531
  reboots
2532
- Removed dependency on debian's patched fping that uses the
2533
  non-standard ``-S`` option
2534
- Now the OS definitions are searched for in multiple, configurable
2535
  paths (easier for distros to package)
2536
- Some changes to the hooks infrastructure (especially the new
2537
  post-configuration update hook)
2538
- Other small bugfixes
2539

    
2540
.. vim: set textwidth=72 syntax=rst :
2541
.. Local Variables:
2542
.. mode: rst
2543
.. fill-column: 72
2544
.. End: